# On the ancestry and evolution of the extinct dire wolf

> An annotated edition of this paper on Fermat's Library, where scientists and readers explain its ideas, equations and history in the margins. Annotated edition: https://fermatslibrary.com/s/on-the-ancestry-and-evolution-of-the-extinct-dire-wolf

- **Authors:** Gregory L. Gedman et al.
- **Published:** 2025
- **Topic:** Biology
- **Original:** https://s3-us-west-1.amazonaws.com/fermatslibrary/uploads%2F1744592394025-jm11qthfy9f-02a44e221e79e30af03a5f8f134e8ad6%2F2025.04.09.647074v1.full.pdf
- **Pages:** 42
- **Annotations:** 4 comments in 4 threads
- **Featured:** Fermat's Library Journal Club
- **Annotated edition:** https://fermatslibrary.com/s/on-the-ancestry-and-evolution-of-the-extinct-dire-wolf

## Annotations

### Annotation 1 (page 1)

Beside the passage: “1 Colossal Biosciences, Dallas, TX 75247, USA”

**Annotation** by Tymor Hamamsy (2025-04-20):

> Colossal Biosciences is an American biotechnology and genetic engineering company working to de-extinct several extinct animals, including the woolly mammoth.
>
> For more: https://en.wikipedia.org/wiki/Colossal_Biosciences

### Annotation 2 (page 2)

Beside the passage: “SUMMARY Dire wolves ( Aenocyon dirus ) are extinct predators of Pleistocene North America. Although phenotypically similar to living wolves ( Canis lupus ), dire wolves have yet to be placed con ﬁdently in the canid family tree. We generated 3.4× and 12.8× paleogenomes from two well-preserved dire wolves dating to > …”

**Annotation** by Tymor Hamamsy (2025-04-20):

> The dire wolf is an extinct species of canine which was native to the Americas during the Late Pleistocene and Early Holocene epochs (125,000–10,000 years ago).
>
> For more: https://en.wikipedia.org/wiki/Dire_wolf

### Annotation 3 (page 3)

Beside the passage: “Dire wolves were morphologically similar to gray wolves ( Canis lupus ) with some exceptions. Dire wolves had larger temporalis muscles than gray wolves, capable of generating a greater bite force 3 . They also had proportionately longer bacula, suggesting sexual competition 4,5 . Dire wolves were comparable in size …”

**Annotation** by Tymor Hamamsy (2025-04-20):

> The Pleistocene (referred to colloquially as the Ice Age) is the geological epoch that lasted from c. 2.58 million to 11,700 years ago, spanning the Earth's most recent period of repeated glaciations.
>
> For more: https://en.wikipedia.org/wiki/Pleistocene

### Annotation 4 (page 4)

Beside the passage: “Resolving the phylogenetic position of dire wolves and assessing evidence for gene ﬂow is dicult given the millions of years of divergence between distinct canid lineages and technical challenges of analyzing ancient DNA. While low coverage paleogenomes can be useful for phylogenetic and demographic inference 23,24 , …”

**Annotation** by Tymor Hamamsy (2025-04-20):

> Paleogenomics is a field of science based on the reconstruction and analysis of genomic information in extinct species. Improved methods for the extraction of ancient DNA (aDNA) from museum artifacts, ice cores, archeological or paleontological sites, and next-generation sequencing technologies have spurred this field. It is now possible to detect genetic drift, ancient population migration and interrelationships, the evolutionary history of extinct plant, animal and Homo species, and identification of phenotypic features across geographic regions.
>
> For more: https://en.wikipedia.org/wiki/Paleogenomics

## Full text

Extracted from the PDF. [Annotation N] marks where each annotation above sits.

#### Page 1

On the ancestry and evolution of the extinct dire wolf

AUTHORS

Gregory L. Gedman

1 *, Kathleen Morrill Pirovich

1 *, Jonas Oppenheimer

1,2

, Chaz Hyseni

1 , Molly

Cassatt-Johnstone

3 , Nicolas Alexandre

1 , William Troy

4 , Chris Chao

1 , Olivier Fedrigo

1 , Savannah

J. Hoyt

5 , Patrick G.S. Grady

1,4,5

, Sam Sacco

2 , William Seligmann

2 , Ayusman Dash

1 , Mithil

Chokshi

1 , Laura Knecht

1 , James B. Papizan

1 , Tyler Miyawaki

1 , Sven Bocklandt

1 , James Kelher

1 ,

Sara Ord

1 , Audrey T. Lin

6,7

, Brandon R. Peecook

8 , Angela Perri

9 , Mikkel-Holger S. Sinding

10

,

Greger Larson

11

, Julie Meachen

12

, Love Dalén

13,14,15

, Bridgett vonHoldt

16

, M Thomas P Gilbert

17,18

,

Christopher E. Mason

19

, Rachel J. O'Neill

5 , Elinor K. Karlsson

20,21

, Brandi L. Cantarel

4 , George R.

R. Martin

22

, George Church

1,20,23

, Ben Lamm

1 , Beth Shapiro

1

*Co-ﬁrst authors

1 Colossal Biosciences, Dallas, TX 75247, USA [Annotation 1]

2 Department of Biomolecular Engineering, University of California, Santa Cruz, Santa Cruz, CA

3 Department of Ecology and Evolutionary Biology, University of California, Santa Cruz, Santa Cruz, CA

4 Form Bio, Dallas, TX 75247, USA

5 Institute for Systems Genomics, University of Connecticut, Storrs, CT, USA

6 Richard Gilder Graduate School, American Museum of Natural History, New York, NY, USA

7 Department of Anthropology, National Museum of Natural History, Smithsonian Institution,

Washington, DC, USA

8 Idaho Museum of Natural History & Department of Biological Sciences, Idaho State University

9 Chronicle Heritage, Phoenix, Arizona

10

Dept of Biology, University of Copenhagen, Denmark

11

Palaeogenomics and Bio-Archaeology Research Network, School of Archaeology, University of

Oxford, Oxford, UK

12

Anatomy Department, Des Moines University

13

Department of Zoology, Stockholm University, Sweden

14

Centre for Palaeogenetics, Stockholm, Sweden

15

Department of Bioinformatics and Genetics, Swedish Museum of Natural History, Stockholm, Sweden

16

Department of Ecology & Evolutionary Biology, Princeton University

17

Center for Evolutionary Hologenomics, The Globe Institute, University of Copenhagen, Denmark

18

University Museum, NTNU, Trondheim, Norway

19

Department of Physiology and Biophysics, Weill Cornell Medicine, New York, NY, USA

20

Broad Institute of MIT and Harvard, Cambridge, MA, USA

21

Genomics and Computational Biology, University of Massachusetts Chan Medical School, Worcester,

MA, USA

22

Fevre River Packet Co.

23

Harvard Medical School, Wyss Institute.

1

.CC-BY-NC-ND 4.0 International licenseavailable under a

was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 2

SUMMARY Dire wolves ( [Annotation 2] Aenocyon dirus ) are extinct predators of Pleistocene North America. Although phenotypically similar to living wolves ( Canis lupus ), dire wolves have yet to be placed con ﬁdently in the canid family tree. We generated 3.4× and 12.8× paleogenomes from two well-preserved dire wolves dating to > 13,000 and > 72,000 years ago, and estimated consensus species trees for these and 10 canid species. Our results revealed that ~2/3 of dire wolf ancestry is derived from a lineage sister to the clade comprising the gray wolf, coyote, and dhole, and the remaining ~1/3 from a lineage near the base of Canini diversity. We identi ﬁed 80 genes evolving under diversifying selection in dire wolves. Our results underscore the power of paleogenomes to resolve long-standing taxonomic questions and contribute to growing evidence of the role of post-speciation gene ﬂow as an evolutionary force.

KEYWORDS evolution, wolf, canid, lineage, admixture, selection, ancient DNA, genomics, paleogenomics, pangenomics

2

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 3

INTRODUCTION The dire wolf (Aenocyon dirus ) ﬁrst appeared in the North American fossil record during the late Middle Pleistocene ~250,000 years before present (ybp) 1 and was the most prominent canid species in continental North America until its extinction ~12,900 ybp (radiocarbon calibrated) 2 . Their large size and morphological distinctiveness, as well as their abundance in fossil deposits at the Rancho La Brea Tar Pits (Los Angeles, California) has established dire wolves as iconic predators of the American Ice Age.

Dire wolves were morphologically similar to gray wolves ( Canis lupus ) with some exceptions. Dire wolves had larger temporalis muscles than gray wolves, capable of generating a greater bite force 3 . They also had proportionately longer bacula, suggesting sexual competition 4,5 . Dire wolves were comparable in size to the largest gray wolves – C. l. occidentalis in the northwestern United States 6 – and exhibited similar variation by latitude compared to gray wolves. Based on dierences in size and geography, dire wolves are ascribed to two subspecies: the smaller ~60 kg A. d. guildayi in the west, and the larger ~68 kg A. d. diru s in the east 7,8 . Dire wolves are believed to have originated in North America 9–11 , but their evolutionary history and taxonomic relationships to other canid species remain uncertain. A recent analysis of ~0.01× to 0.23× coverage dire wolf genomes concluded that the dire wolf lineage ﬁrst diverged from wolf-like canids, including gray wolves ( C. lupus ), coyotes ( C. latrans ), African golden wolves ( C. anthus ), golden jackals ( C. aureus ), Ethiopian wolves ( C. simensis ), dholes ( Cuon alpinus ), African wild dogs ( Lycaon pictus ), side-striped jackals ( Lupulella adustus ), and black-backed jackals ( L. mesomelas ), during the late Miocene, ~6 million years ago (Mya) 12 . This timing coincided with the ﬁrst appearance in the Eurasian and African fossil records of the genus Eucyon, which is the lineage from which extant wolf-like canid species probably descended 13 . During the Pleistocene ice ages, shifting habitats and the ephemeral emergence of the Bering Land Bridge provided repeated opportunities for wolf-like lineages to disperse in and [Annotation 3] out of continental North America. Coyotes ﬁrst appear in the North American fossil record 1-2 Mya 14,15 , and gray wolves are present by the late Middle Pleistocene 9,16 , although current North American gray wolf lineage entered North America from Eurasia around 25,000 years ago 17–19 . Despite the geographic and temporal overlap between the ranges of dire wolves, coyotes, and gray wolves, no evidence of admixture between dire wolves and either lineage has been identiﬁed 12 . Since admixture between closely related canid species is common 20–22 , the absence of gene ﬂow between dire wolves and sympatric canids suggests temporal, geographic, reproductive, or behavioral isolation.

Although Perri et al. 12 conﬁrmed that dire wolves are a distinct evolutionary lineage, a high frequency of phylogenetic incongruence among estimated gene trees led to uncertainties about their early evolutionary history. About 50% of reported gene trees supported a topology in which dire wolves formed the earliest-diverging lineage of the wolf-like canid clade (Canina). Of the remaining gene trees, the side-striped and black-backed jackals were found to be an outgroup to a clade comprising dire wolves and other wolf-like canids 12 . Divergence time estimates suggest the dire wolf lineage emerged after the divergence of the South American Cerdocyonina subtribe 5-4 Mya 12,23 and before the divergence of the common ancestors of African jackals ( Lupulella spp.) 7.6-3.5 My and wolf-like canids 8.5-4.0 Mya (Canina subtribe). Incomplete lineage sorting 24 resulting from

3

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 4

such rapid diversiﬁcation could be one driver of this incongruence. In addition, while at least one admixture event between dire wolves and the common ancestor of dholes, coyotes, and wolves was detected, additional gene ﬂow from unsampled populations, like the South American canids, could not be excluded.

Resolving the phylogenetic position of dire wolves and assessing evidence for gene ﬂow is dicult given the millions of years of divergence between distinct canid lineages and technical challenges of analyzing ancient DNA. While low coverage paleogenomes can be useful for phylogenetic and demographic inference 23,24 , ancient DNA is susceptible to DNA damage, deamination, and reference bias in mapping, in which divergent reads fail to map and similarity to the reference genome is artiﬁcially inﬂated 25 . This bias results in an overestimation of sequence similarity between the ancient genome and the mapped reference, which can impact phylogenetic reconstruction and reconstructions of population history 26,27 . This bias is particularly problematic when the reference genome is evolutionarily distant from the paleogenome 28,29 , as is the case for the dire wolf. Several factors contributed to the eventual extinction of dire wolves 30,2,31 . The extinction of megafaunal prey species may have led to nutritional stress among dire wolves. The fossil record shows evidence of shrinking skull sizes during the Last Glacial Maximum 32 , which can result from changes in diet and lifestyle. The loss of dire wolf prey and habitat from human-induced ﬁres during the Late Pleistocene has been one proposed driver of these changes leading to their decline 2 . Reduced population sizes may also have led to diversity loss and the accumulation of deleterious genetic variation due to increased rates of inbreeding. In support of this hypothesis, ossi ﬁcation disorders are observed in limb bones dating to several thousand years prior to dire wolf extinction 31 . To better understand the origin, evolution, and extinction of the dire wolf, we leveraged improved paleogenomics approaches and recovered additional ancient DNA from two [Annotation 4] well-preserved dire wolves 12 : DireSP, an isolated incisor (Cincinnati Museum Center VP1737) recovered from Sheriden Pit, Ohio, estimated to be ~13,000 years old based on calibrated radiocarbon dates from other bone material from the site 33 . and DireGB, a complete isolated skull (Idaho Museum of Natural History, IMNH 48001/52) from Gigantobison Bay, a beach around the American Falls Reservoir, Idaho, dated to at least 72±14 thousand years ago based on the age of volcanic material from a nearby stratum 34 . We generated paleogenomes for both specimens and performed phylogenetic, admixture, and selection analyses that provide new insights into the evolution and adaptation of the dire wolf to its unique niche.

RESULTS We generated paleogenomes with 3.4× depth of coverage of DireSP and 12.8× coverage of DireGB. After ﬁltering, trimming, and merging overlapping reads, we retained 2.8 billion unique sequence reads for DireSP and 33.3 billion unique sequence reads for DireGB (Table S1). We mapped the reads using BWA aln against the Greenland gray wolf chromosome-level (including Y) genome assembly (GenBank GCA_905319855.2 aka “canLor”) 35 (Table S2). After mapping, 6% of the genome was covered >5× for DireSP and 82% for DireGB (Fig. 1A-B, Table S3). Previously, ancient DNA isolated from these same individuals yielded 0.2× (DireSP) and 0.3× (DireGB) coverage 12 (Table S3). We found patterns of DNA damage expected for single-stranded libraries (5' C>T: DireGB = 30.2%; DireSP=26.6%, Fig. S1). Mapping to the gray

4

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 5

wolf reference sex chromosomes versus autosomes yielded Rx ratios 36,37 indicating that both individuals were male (95% conﬁdence intervals: DireSP=0.539-0.554, DireGB=0.538-0.554).

We next used an iterative process to reconstruct a progressively less divergent nuclear paleogenome reference (“aenDir”) for mapping dire wolf ancient DNA sequencing libraries. After six iterations, we identi ﬁed 14,569,091 changes (13,665,448 single nucleotide variants, 388,514 insertions, and 515,129 deletions) from the seed reference assembly (gray wolf). We then called a pseudo-haploid, variant-based consensus sequence that represented the highest depth alleles. In parallel, we used established methods 28,38–40 to reconstruct the mitochondrial paleogenome at high coverage (average depth = 158.4×). Merging both nuclear and mitochondrial reconstructed paleogenomes resulted in a ﬁnal reference assembly with a hard-masked N content of 16.84% (Table S4).

Figure 1. Iterative reconstruction of ancient dire wolf nuclear paleogenome improves whole genome and exome mappability. (A) Mean read depth of coverage by DireGB libraries (MQ>25) mapped over 1 Mb windows from the reference assembly of A. dirus reconstructed by iterative mapping and polishing. Inset displays mean read depth over the reconstructed full-length mitochondria sequence split by 1 Kb bins. (B) Cumulative coverage plot of DireGB (purple) and DireSP (green) libraries mapped to divergent gray wolf reference “canLor” (dotted line) and iterative dire wolf reference “aenDir” (solid line). Proportion of reference genome (i.e., breadth of coverage) at 5× and average depth for DireGB against the dire wolf reference are highlighted. (C) Average coverage dierence over annotated orthologous genes (n=13,230) relative to dire wolf (positive) and gray wolf (negative). The majority of genes exhibited a boost in average coverage (dotted line) when using the reconstructed dire wolf paleogenome assembly in both DireGB (77% +3.3×) and DireSP (97% +2.1×).

5

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 6

Mapping ancient DNA libraries to the reconstructed paleogenome reference (aenDir) increased coverage both at the whole genome level and at the exome level (Fig. 1A-B). Compared with the initial mapping to the Greenland gray wolf genome reference, mapping to aenDir resulted in an increase in average coverage of 2.8× for DireGB and 1.5× for DireSP libraries. Following annotation of the reconstructed paleogenome (aenDir), we found that the majority of protein-coding genes (71%) were retained compared to the gray wolf reference (canLor), with most missing genes due to masking of low coverage regions (Figs. S2,3). The depth of coverage in exons improved by ~2-fold for both DireSP and DireGB when mapped against the reconstructed dire wolf paleogenome (Fig. 1C, Table S5).

Mapping against the reconstructed paleogenome (aenDir) also reduced the impact of reference bias (Fig. 2A, Table S6). We found a signi ﬁcant decrease in allele sharing for the DireGB consensus called against the reconstructed dire wolf (aenDir) paleogenome compared to the gray wolf (canLor) genome, estimated using the D -statistic D (seed reference Canis lupus , non-seed reference Canis lupus ; consensus Aenocyon dirus , outgroup Vulpes vulpes ) ( D for aenDir=-0.152, Z -score=-16.01; D for canLor=-0.521, Z -score=-76.51). In a complementary test, we assessed the identity-by-state (IBS) for DireSP reads mapped against the reconstructed aenDir paleogenome (Fig. 2B). We found that DireSP reads mapped to our reconstructed dire wolf genome exhibited higher IBS (i.e. greater sequence identity) compared to the gray wolf seed reference across all read lengths, especially for reads < 35 bp, suggesting the presence of residual reference bias, predominantly in shorter reads.

Dire wolf paleogenomes were unexpectedly more similar to extant Canina genomes than anticipated given a 5.8 Mya divergence time 12 , a mutation rate calibrated for Canis species 41 , and a wide range of generation times 42 for priors (Fig. S4). We computed sequence divergence on ﬁxed autosomal transversions at sites covered at 5× relative to the total non-N sequence length of the reconstructed dire wolf paleogenome reference (2,310,286,013 bp). We found the dire wolf paleogenomes less diverged (on average, 0.09% divergent) than expected based on prior parameters (0.6% to 0.9% sequence divergence), and estimated sequence similarities of 99.91% to jackals and 99.94% to non-seed reference gray wolves at either end of wolf-like canid diversity (Fig. S4A). As residual reference bias in mapping could artiﬁcially inﬂate sequence similarity, we also assessed the distribution of unique 32-mers from the contaminationﬁltered sequence reads using mash for metagenome distance estimation 43 . We found that dire wolves exhibited equally low 32-mer diversity compared to gray wolf and jackal species (~0.01), with greater distance to red and gray foxes (0.02 to 0.03) (Fig. S4B, Table S7).

6

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 7

Our estimated mitochondrial and nuclear phylogenies were similar to those published previously (Figs. 3, S5-9, Tables S8-10). The mitochondrial phylogeny showed the same mitonuclear discordance that is known to characterize the canid lineage 22 . Our nuclear phylogeny, which was based on 19.5Mb of non-coding dire wolf sequence at least 1Mb away from protein-coding genes and <25% missingness on average (Figs. S6,7), placed dire wolves into a monophyletic clade that diverges prior to the split between the black-backed jackal and wolf-like canids (Figs. 3, S8). However, this placement was only supported by 44% of input loci, while 39% of loci supported a closer relationship between dire wolves and non-Lupulellan wolf-like canids (Fig. 3). Divergence time estimates (Fig. S9, Table S10)

7

Figure 2. Iterative process to reconstruct the extinct dire wolf paleogenome sequence reduces reference bias. (A) Mapping to the reconstructed dire wolf paleogenome (aenDir) reduced reference bias 0.7 to 0.8 fold compared to mapping to the gray wolf reference genome (canLor). Residual reference bias was assessed using the ABBA/BABA test D statistic, which measures derived allele sharing of the DireGB consensus genome called against either the seed gray wolf reference genome (canLor) or reconstructed dire wolf paleogenome (aenDir) reference. Ancestral (A) and derived (D) alleles were localized by aligning either of the query ( A. dirus in red, Q) consensus DireGB sequences (canLor- or aenDir-called), seed reference Greenland gray wolf ( C.lupus orion in light blue, P R ) and non-seed reference North American gray wolf ( C. lupus in dark blue, P N ) assemblies, and outgroup red fox assembly ( V. vulpes in black, O) to a common reference genomic coordinates from a gray fox ( U.cinereoargenteus ) assembly. (B) Average identity-by-state (IBS) by read length for DireSP reads mapped to either the DireGB-based iteratively reconstructed paleogenome (dark purple) or divergent gray wolf (light purple) reference assemblies. IBS is deﬁned as (matches - mismatch/aligned read length). Averages were calculated from a 500k read sample of the total alignment with an average of 5.7k (gray wolf) and 6.3k (dire wolf) reads per size bin.

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 8

suggested that the dire wolf diverged from other canids ~4.5 million years ago (95% CI=5.84-2.89 Mya) and the jackals diverged ~4.08 million years ago (95% CI=5.34-2.58 Mya).

Figure 3. The dire wolf lineage is distinct from jackals and other wolf-like canids. Time-calibrated nuclear phylogenetic relationship of extinct dire wolves to extant canids using MCMCtree from 25 kb loci trees (n=924) using IQTree and species tree consensus using ASTRAL. Dire wolves (red) diverge prior to the split of black-backed jackals (yellow) and wolf-like canids (blue). Root ages are mean divergence time estimates in millions of years (Mya). Node distributions visualize conﬁdence in placement. Node distributions with fossil calibrations are shown in purple. Inset details the alternative topologies for the dire wolf node, along with the proportion of loci supporting each alternate.

The discordance among estimated gene tree topologies appears to arise from admixed ancestry in dire wolves (Fig. 4). Dire wolves shared an excess of derived alleles with gray wolves relative to other canids, including dholes (11.9 ≤ Z ≤ 15.3), African wild dogs (61.9 ≤ Z ≤ 65.4), African jackals (92.3 ≤ Z ≤ 99.6), and South American canids (173.4 ≤ Z ≤ 183.7) (Fig. 4A). We did not detect similar anity between dire wolves and gray wolves when comparing to wolf-like canids that are more closely related gray wolves, including coyotes and red wolves ( |Z| ≤ 2.4) (Fig. 4A). We also observed that, relative to dire wolves, African jackals (66.0 ≤ Z ≤ 72.4) and other wolf–like canids ( Z ≥ 92.7) share more derived alleles with gray wolves (Fig. 4B). Together, these patterns suggest that dire wolves had a mixture of ancestries from two distinct lineages, with one diverging prior to African jackals and another stemming before gray wolves and coyotes. Allele sharing between dire wolves and South American canids relative to extant wolf-like canids (-7.2 ≤ Z ≤ -2.3) is consistent with the deeper ancestry component falling within Cerdocyonina diversity (Fig. 4B). Both dire wolves

8

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 9

formed a clade relative to all individual coyotes and gray wolves (Fig. S10), implying that there was limited gene ﬂow between dire wolves and speciﬁc Canis lineages following their divergence.

Admixture graph models of canid population history identiﬁed both wolf-like canid and deep ancestry components in the dire wolf lineage (Fig. 4C). We applied a heuristic algorithm 44 to ﬁnd optimal admixture graphs modeling the relationships between seven canids (dire wolves, Eurasian gray wolves, coyotes, dholes, African wild dogs, maned wolves, using red foxes as an outgroup). No model permitting zero or one admixture event was supported by the data; however, the bestﬁtting model featuring two admixture events had statistical support (worst residual of Z =2.6) (Fig. 4C). This model found that dire wolves received a large portion of their ancestry (~39%) from a lineage that groups with South American canids, with the remaining ancestry (~61%) attributed to a sister lineage to dholes, coyotes, and gray wolves. This wolf-like ancestor may have been the source of the phylogenetically-discordant dire wolf mitochondrial genome, which groups within wolf-like canid diversity. When analyzed separately, both dire wolf individuals produced similar optimal graphs (Fig. S13), suggesting that admixture was a species-wide feature of dire wolf population history. Graphs that include both an additional outgroup (gray foxes) and additional South American canid species con ﬁrmed that the divergent portion of dire wolf ancestry falls near the base of Canini diversity (Fig. S13).

The phylogenetic placement of the wolf-like component in dire wolves is remarkably similar to a proposed unsampled canid lineage that contributed ancestry to the common ancestor of coyotes and wolves 45 . The best-ﬁtting model similarly features gene ﬂow from dire wolves into wolves and coyotes, which could be the source of this unsampled ancestry. However, scenarios in which dholes, rather than coyotes and wolves, are admixed also provide similar ﬁts to the data (Table S11) and we observed that relationships among extant wolf-like canids can be modeled without requiring this unsampled admixture (Fig. S12). As suggested by evidence of excess derived allele sharing between dire wolves and South American canids, a similar model which instead places the minor ancestry component of dire wolves outside extant Canini diversity instead of on the South American canid branch provides a worse ﬁt for the data (worst residual of Z =5.25). Models of canid population history that include additional extant canid lineages highlight other known and cryptic admixture events and identify the gene ﬂow event leading to dire wolves (Fig. S14). We found mixed support for introgression from the dhole lineage into the African wild dog lineage, which could re ﬂect cryptic admixture 45,46 , and consistent support for other previously detected admixture events 20,45,47,48 . These ﬁndings reinforced that this approach is sensitive enough to detect hybridization across diverse canids (Fig. S18), and does not appear to be impacted by choice of reference genome (Fig. S13).

The ubiquity of gene ﬂow among canids, especially among diverged wolf-like canid species 45–49 , poses challenges for inferring the precise placement of admixture events among these taxa. Our goal was to infer the simplest models which described the data, both for the minimal (Fig. 4C) and expanded sets of canids (Fig. S14), which each provided evidence for the dual ancestry of dire wolves. However, given extensive hybridization among canids, the true population history may be more complex. More complicated models, such as those featuring three admixture events in the minimal canid species set, also largely suggest hybridization in dire wolves, though a much wider range of graphs ﬁt the data, including many which have other lineages, rather than dire wolves, as admixed (Table S11). Challenges

9

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 10

arising from mapping sequence data diverged taxa also complicate inferences of canid population history. We observed several biases across alignment and genotype calling strategies (Fig. S10, S19). The Consensify-style pseudohaploid genotype calls using pangenome alignments presented here appeared to minimize these biases (Fig. S13, S19), though do require greater agreement between reads, which has the potential to amplify biases arising in mapping.

Figure 4. Investigating gene ﬂow between dire wolves and other canid lineages. (A) D-statistics of the form D (gray wolf, other Canini group; dire wolf, red fox), investigating allele sharing between dire wolves and gray wolves, relative to other species and subspecies within Canini. (B) D -statistics of the form D (gray wolf, dire wolf; other Canini group, red fox), investigating allele sharing between dire wolves and other populations within Canini, relative to gray wolves. Bars depict 3 standard errors in (A) and (B). Each point is a comparison with a dierent individual gray wolf, with points colored by the Canini clade being of the non-gray wolf population in the comparison. (C) Best-ﬁtting admixture graph with two admixture events modeling the relationship between dire wolves (n=2); wolf-like canids: Eurasian wolves (n=2); coyotes, (n=4), dholes (n=2), African wild dogs (n=3); and maned wolf (n=2); with red fox (n=2) as the outgroup. Additional graph ﬁts are summarized in Table S11.

Comparative sequence analysis detected molecular signatures for diversifying and adaptive evolution in dire wolf protein-coding genes (Fig. 5). We used the HyPhy framework 50–52 to identify genes with signatures of positive selection among dire wolves and 10 other canid taxa, including episodic diversifying selection in dire wolves contrasted to other canids. We identiﬁed 49 genes under signiﬁcant (BUSTED p < 0.05) episodic positive selection and another 31 genes under non-episodic positive selection in the dire wolf branch (free-ratio dN/dS > 2.0) (Fig. 5, Table S12). None of these genes exhibited pervasive positive selection over the canid tree (one-ratio dN/dS < 2.0), although dire wolves may share positive selection at a site level with other branches. Rarely, we found that the dire wolf codon matched a codon

10

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 11

found in another branch. For example, the 505th codon of NCAPG (MEME p = 3.31e-06) had two branches under positive selection: the root codon encodes serine (TCT) but both the dire wolf and Canis branches, and not other wolf-like canids, encode asparagine (AAT). Likewise, the 794th codon of WNK4 (MEME p = 0.00513), encoded asparagine (AAC) in both dire wolf and Lycaon branches rather than the root alanine (GCC). We did not ﬁnd extensive functional enrichment for gene sets across the 80 genes under diversifying positive selection in the dire wolf branch. Functional domain analysis 53 and protein language modeling 54 identiﬁed 42 dire wolf codons in 22 genes under episodic diversifying selection that intersect protein functional domains which could be impacted by amino acid substitution (Table S13).

Figure 5. Comparative analysis of nuclear protein-coding gene sequences supported by dire wolf ancient DNA and 10 other canid genomes inform pathways under episodic or pervasive diversifying positive selection that may shape species adaptations and phenotypes. The -log10 HyPhy BUSTED p -values for episodic positive selection in the dire wolf branch plotted (y-axis) over the canid tree-wide, one-ratio dN/dS ratio maximum likelihood estimates (x-axis) for 5,984 genes where high one-ratio dN/dS implies pervasive positive selection across the canid lineage. We ﬁnd 31 genes exhibiting branch-speciﬁc, free-ratio dN/dS > 2 (orange triangles), 28 genes signiﬁcant for branch-speciﬁc, episodic selection in BUSTED test p < 0.05 (green squares), or 21 genes satisfying both conditions (purple diamonds). Annotated genes without gene symbols of orthologs identiﬁed are labeled with “FB:”.

DISCUSSION Dire wolf evolutionary history

11

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 12

The Late Pleistocene dire wolf is descended from a canid lineage that emerged following admixture between two lineages: a lineage along the branch that gave rise to the South American canids (Cerdocyonina subtribe) after they diverged from African jackals and other wolf-like canids (Canina subtribe), and a more recently derived wolf-like canid lineage that diverged after the emergence of the African wild dog ( Lycaon ) lineage (Figs. 3, 4C). This admixed ancestry explains the discordance in nuclear gene phylogenies observed both here and previously reported in Perri et al. (2021).

Given fossil evidence that dire wolves were only present in North America, the ancestral components of the dire wolf lineage might be attributed to other canids in the North American fossil record. The oldest wolf-like fossils in the North American fossil record are from the tribe Canini 10 , including the extinct genus Eucyon (10.3-3.6 Mya) , the proposed progenitor of Canis and Lupulella , and the earliest ancestors of the Cerdocyonina subtribe, which dispersed to South America 55–57 . It would be plausible, given an estimated divergence timing straddling Late Miocene and Pliocene, that the major ancestry component of the dire wolf is from an early North American Eucyon species, but our admixture graph modeling supports instead a major ancestor diverging after the Lycaon lineage (African wild dogs), which would be post Eucyon expansion outside of North America . Eucyon expanded into Eurasia and Africa during the Late Miocene, where it gave rise to Lupulella , the African jackals 57–61 , as well as Canis and the related genera Lycaon and Cuon 46 . From the late Miocene through the Pleistocene, various Canis species expanded back into North America, supplanting or admixing with local populations 13 . Canis lepophagus expanded into the Americas during the late Miocene to early Pleistocene 10,62 , for example, followed by C. armbrusteri and C. edwardii in the Pliocene and Pleistocene. The wolf-like lineage that contributes majority ancestry to dire wolves could be one of these earlier expanding lineages.

We note that the deep divergence (~10 Mya) among the canid lineages examined here makes it challenging to examine gene ﬂow across these species while accounting for reference bias, particularly with highly fragmented ancient DNA. While our iterative mapping approach mitigated reference bias in our alignment of dire wolf sequences, it is possible that our more fragmented dire wolf reference reduced mapping performance and ampliﬁed other biases in the alignment of short reads from non-dire wolf canids. Thus, we complemented analyses of canid population history with a pangenome-mapping approach which mitigated these biases. Additionally, the prevalence of interspeciﬁc gene ﬂow among canids makes it unlikely that any particular model will fully capture the underlying population history relating these species 47 . However, our conclusion regarding the dual ancestry of dire wolves is robustly supported across dierent models and approaches.

Our results underscore the role of post-divergence gene ﬂow as an evolutionary force shaping species across the tree of life, and the potential for lineages to be forged through hybrid speciation – as uncovered in the Columbian mammoth ( Mammuthus columbi ), the same megafauna that roamed alongside dire wolves 63,64 . For example, excess basal ancestry in wolves and coyotes relative to other wolf-like canids was previously attributed to gene ﬂow with an unknown wolf-like canid lineage that emerged following the divergence of African wild dogs 65 , raising the possibility that the same unknown lineage interbred with both groups. Indeed, our results suggest that the wolf-like lineage found within dire wolves also contributed ancestry to wolves and coyotes, possibly mediated through gene ﬂow with dire wolves. However, we infer that gene ﬂow between diverged canid lineages was common, and that the relationships within the wolf-like canids can be modeled without invoking this

12

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 13

unknown lineage. Our admixture graph analyses do, however, uncover previously described episodes of gene ﬂow between canids, resulting in admixed ancestries for species including eastern coyotes 48 , red wolves 47 , golden wolves 20,45 , and dholes 46 .

Evolutionary adaptations of dire wolves

We identiﬁed several genes in the dire wolf genome that may have contributed to one of their primary evolutionary features: their large builds. NCAPG, a gene associated with body size, was the most signi ﬁcant gene for episodic diversifying selection in the dire wolf branch. Yet, dire wolves and gray wolves exhibit unusually low rates of synonymous substitutions in NCAPG (both dS= 4e-10) – magnitudes lower than other canids (all dS > 0.001) – and together share a serine-to-asparagine substitution under site-speci ﬁc positive selection (MEME p = 3.31e-06). This degree of sequence constraint could suggest strong purifying selection 66–69 , perhaps upon non-coding regulatory elements inﬂuencing expression of adjacent gene LCORL , just 200 kilobases away. LCORL is an allosteric regulator of the polycomb repressor complex (PRC2) 70 , and haplotypes spanning NCAPG and LCORL have been strongly associated with growth traits in domestic dogs 71 , goats 72 , sheep 73 , horses 74 , and cattle 75,76 . Expression levels of NCAPG and LCORL have inverse relationships on growth: positively correlated for NCAPG levels 77 and negatively correlated for LCORL levels 74 . We hypothesize that in both dire wolves and gray wolves, purifying selection has maintained ancestral regulatory elements at the NCAPG locus, while in other wolf-like canids of smaller body sizes, diversifying selection altered co-regulation of these genes.

We also found several genes under episodic diversifying selection in dire wolves annotated as aecting germ cell development, gamete function, and reproductive tissue structure, which may contribute to the reproductive biology of dire wolves distinct from wolf-like canids 4 . The genes B4GALNT1 , ROS1 , TP73 , AKAP3 , NUTM1 , and ABCG5 are all associated with “male infertility” in the Mammalian Phenotype Ontology (MP:0001925). While most of these fertility-related genes have not been implicated in reproductive structures, ROS1 regulates the dierentiation of male epididymis 78 . PDGFA , another gene under positive selection in dire wolves, has been shown as integral to epididymal morphology in mice 79 . None overlap the few genes associated with mammalian baculum size, however 80 , meaning that genes shaping divergent reproductive biology in dire wolves are still a mystery.

An iterative approach to nuclear genome inference We showed that iterative mapping combined with conservative consensus calling can help mitigate the impacts of reference bias and DNA damage in paleogenomics. To generate the dire wolf paleogenome, we mapped reads to the Greenland gray wolf as a seed reference, called consensus from aligned dire wolf reads, then re-mapped reads to this new consensus. We then repeated this process until no additional reads were added. Our intent was to estimate a conservative assembly, and thus we chose to hard-mask bases (i.e., Ns) not supported by >5× coverage, resulting in a ﬁnal reference assembly comprising only the most abundant alleles from mapped dire wolf reads. Methods like EAGER use a similar conservative consensus calling approach when mapping to the nuclear genome, but employ only a single round of mapping and thus do not capture the maximum number of reads 81 . Iterative assembly tools are used in ancient DNA research, but mostly for mitochondrial genome assembly (e.g., MIA 82 ) as they do not scale to whole nuclear genomes 83 . Iterative mapping was used to reconstruct the Denisovan genome, where Meyer et al. (2012) 38 performed two

13

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 14

rounds of iterative mapping and then ﬁlled in low coverage regions with human sequence, which is a source of reference bias. Given the improvements that we observed with multiple rounds of iterative mapping, we recommend a similar approach, in particular when no close reference genome is available.

One concern with an iterative mapping approach is that the process could actually exacerbate reference bias rather than reduce it. If the seed reference is too divergent, for example, only the most similar ancient reads will map, potentially masking informative sites and compounding bias. We approximated reference bias using D -statistics 38,84 by assessing derived allele sharing with distinct gray wolf subspecies, and found no evidence of compounding bias. In fact, we observed a reduction in bias towards the seed reference Greenland gray wolf genome. Since shorter reads have been shown to be more impacted by reference bias 40,85 , we performed a secondary assessment of identity by state (IBS) by read length from an unrelated dire wolf individual (DireSP). While residual reference bias remains, particularly for shorter reads, we observed higher IBS on average for all DireSP reads regardless of length, suggesting we successfully generated a less divergent reference genome that facilitated better mapping of ancient reads.

While our approach estimated a paleogenome that was more similar to the dire wolf relative to the seed reference, steps could be taken to improve these genomes further. For example, while our approach can identify and incorporate SNPs and small insertions or deletions into the ﬁnal reference, it cannot account for larger structural variants speciﬁc to the dire wolf. The reconstruction of large structural variants is unresolved for short read data sets that are typical of paleogenomes 86 and, given our average read length of 69 bp across all samples, is not achievable for the dire wolf. Nevertheless, by iteratively estimating a paleogenome reference, we could recover more reads that would have otherwise missed mapping to a divergent reference genome without those small sequence dierences.

An alternative – or, perhaps, complementary – approach to mitigating reference bias is to construct a pangenome from multiple divergent species for initial mapping 87,88 , which allows for capture of highly divergent reads and, potentially, larger structural variants. The bene ﬁts of a pangenome approach were demonstrated by Martiniano and colleagues (2020) 89 , who saw a reduction of reference bias in mapping simulated ancient reads to the 1000 Genomes Project variational graph. By mapping real reads from ancient humans to this pangenome, they could capture sequence variation missed by initial mapping to a single reference without a loss of mapping sensitivity. However, divergent dire wolf haplotypes not represented by the available paths in the pangenome will still be aected by reference bias. Nevertheless, cost-ecient optimizations that enable a combination of iterative mapping and polishing approach will provide additional improvements to estimating increasingly accurate paleogenomes from short read data sets.

These data provide a comprehensive overview of dire wolf genomics, evolution, and adaptations, oering insights into this iconic species. The reconstructed reference paleogenome provides a resource for the broader research community and a roadmap for future studies aiming to leverage high-coverage ancient DNA. The genes identiﬁed under selection in dire wolves oer insights into possible molecular mechanisms for this species' role as a predator of extinct megafauna. This work furthers our understanding of the dire wolf and their place in the broader patterns of canid evolution.

14

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 15

STAR METHODS

Published ancient DNA sequencing data We downloaded ancient DNA sequencing reads from dire wolf individuals previously sampled, including the Sheriden Pit individual (DireSP, BioSample accession SAMEA5559854) and the Gigantobison Bay individual (DireGB, BioSample accession SAMEA5559855) and three other individuals from American Falls Reservoir, Idaho (DireAFR, BioSample accession SAMEA5559852), Natural Trap Cave, Wyoming (DireNTC, BioSample accession SAMEA5559856), and Guy Wilson Cave, Tennessee (DireGWC, BioSample accession SAMEA5559853) publicly released under BioProject accession PRJEB31639 (Table S1).

Provenance and sampling of dire wolf individuals We obtained permission to collect new samples from two dire wolf individuals that previously yielded authentic dire wolf mitochondrial and nuclear DNA 12 . The ﬁrst dire wolf, referred to as “DireSP” (Cincinnati Museum Center; VP1737; ACAD1735; DireSP), is from Sheriden Pit site in Indian Trail Caverns, Ohio, and an incisor tooth previously sampled at the root (BioSample accession SAMEA5559854). We sampled an additional incisor from this individual for the present study. The second dire wolf, referred to as “DireGB” (Idaho Museum of Natural History; IMNH 48001/52; ACAD18742; DireGB) is from Gigantobison Bay site at American Falls Reservoir, Idaho, and presents a complete cranium with intact teeth. The left petrosal was previously sampled (BioSample accession SAMEA5559855) and the right petrous was sampled for this study. The Sheriden Pit individual could have been deposited before 13,000 years ago, based on calibrated radiocarbon dates from other bone material derived from the same layer 33 , but has not been radiometrically dated, and the Gigantobison Bay individual is probably older than 72,000 years, based on the age of volcanic material from a nearby stratum 34 .

Sample preparation and ancient DNA extraction For the Sheriden Pit individual (BioSample accession SAMN46779696, we extracted ancient DNA from an incisor (Cincinnati Museum Center VP1737). We used a Dremel with drill attachment to obtain bone powder from the interior of the root. We rinsed the obtained bone powder in a 0.5% bleach solution for ﬁfteen minutes followed by three water washes before proceeding with extraction. For the Gigantobison Bay individual (BioSample accession SAMN46779697), we extracted ancient DNA from the incus bone recovered from the right petrous of a complete skull (IMNH 48001/52). Ossicles are an optimal source of DNA preservation 90,91 . We rinsed the whole incus in a 0.5% bleach solution, followed by three water washes. We extracted DNA for both individuals using the Buer D option, a protocol optimized for recovering short, degraded molecules 92 . For the Sheriden Pit sample, we extracted DNA from two 5 mg aliquots of bone powder following bleach pre-treatment. From these two DNA extracts we generated three single stranded libraries each 93 . Following the initial round of extractions we observed a non-dimineralized bone powder pellet in one of the

15

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 16

two lysate tubes and performed a re-digestion of this bone powder following Rohland et al. (2018) 92 . From the DNA isolated after this re-digestion we generated two more single-stranded libraries 93 . For the Gigantobison Bay individual, we extracted DNA from the whole incus, which was completely dissolved following overnight lysis. We then prepared single-stranded libraries from the extracts as per Kapp et al. (2021) 93 . We generated a single extract from the Gigantobison Bay ossicle, from which we made six libraries. All pre-PCR work, including sample preparation, DNA extraction, library construction, and quality control sequencing, was carried out in a dedicated ancient DNA lab at the University of California, Santa Cruz.

Massively parallel paired-end short DNA read sequencing We prepared a total of nine Illumina libraries from the sample of the Sheriden Pit dire wolf (DireSP) and six Illumina libraries from the sample of the Gigantobison Bay dire wolf (DireGB). We sequenced all libraries on a Nova-seq 6000 using version 1.5 chemistry, loading 0.3nM ﬁnal library concentration. For the DireSP sample, we obtained 1,320gb (R1+R2) of raw sequencing data or 13.2 billion total paired end reads (R1+R2). For the DireGB sample, we obtained 7,814gb (R1+R2) or 78 billion total paired end reads (R1+R2). We report the total number of raw sequencing reads generated per run, library, and individual in Table S1. For subsequent steps of parallelized processing, we split sequencing reads into batches of 50,000,000 reads.

Sequence read pre-processing and quality control We trimmed raw sequencing reads using fastp 94 ( i ) to remove adapter sequence, low quality ends of reads, and reads less than 35 nucleotides, ( ii ) merge and correct overlapping reads with a minimum of 15 nucleotide overlap, and no more than three dierences, and ( iii ) trimmed homopolymer G and X sequences at the 3’ end (-q 25 -l 35 -m --overlap_len_require 15 --overlap_di_limit 3 -c -g - x). We classiﬁed sequence contaminants using Kraken 2 95 and a conﬁdence of 0.8 against the pre-compiled 8 GB database MiniKraken DB_8GB constructed from complete bacterial, archaeal, and viral genomes in RefSeq (2017-10-18). We used combine_kreports.py from Kraken Tools 96 to merge all metagenomic classiﬁcation reports generated for split read batches. We report statistics on sequencing reads before and after processing for both published and novel ancient DNA sequencing libraries in Table S1.

Reference genome assemblies and annotations We selected the Greenland gray wolf ( Canis lupus ) chromosome-level (including Y) genome assembly (GenBank GCA_905319855.2) 35 as our primary single-reference genome assembly of a modern canid species, and utilized additional canid genome assemblies 97–114 throughout this study as referenced in Table S2. We surmised that the Greenland gray wolf would serve as a suitable scaold reference due to the geographic isolation of Greenland wolves relative to other gray wolves 115 , the fully assembled chromosomes including Y and high contig N50 relative to other canid genome assemblies (Table S2), and to previously inferred phylogenies that suggest it is no more distantly related to dire wolves than any other extant members of the Canina subtribe.

16

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 17

Read mapping to single reference genomes We utilized two primary mapping strategies for the majority of the analyses in this study. Strategy 1 was a CPU-based accelerated Sentieon® implementation of the BWA `aln` algorithm 116 and seed lengths of `-l 16,500` nucleotides, maximum edit distances set by a fraction `-n 0.01` of missing alignments, and `-o 2` maximum gap opens 117 . Strategy 2 was a GPU-accelerated NVIDIA® Clara™ Parabricks `fq2bam` implementation of the BWA `mem` algorithm with settings for short reads (`-M -k 19 -r 2.5`).

Similarly, when mapping ancient reads, we used two strategies for removing duplicate reads and assessing DNA damage. First in the “standard” strategy, we used the PALEOMIX 118 `rmdup_collapsed` function and duplicate removal using SAMtools rmdup. We performed base recalibration with mapDamage 119 to rescale base quality scores in consideration of 12 nucleotides from both 5’ and 3’ ends of sequencing reads (`--rescale-length-5p 12` and `--rescale-length-3p 12`). We pro ﬁled Ancient DNA damage patterns using DamageProﬁler 120 . The second strategy was “stringent,” where we merged each sequencing run across libraries and marked duplicates again using `MarkDuplicates` from the Picard toolkit (broadinstitute.github.io/picard) to remove any remaining duplicate reads at the library level. After additional duplicate removal, we ran mapDamage again to assess overall DNA composition, read length distributions, fragmentation and misincorporation rates and C-to-T and G-to-A transition frequencies at relative positives across reads.

We assessed alignment statistics using samtools `idxstats` and `ﬂagstat`, mosdepth 121 , and QualiMap 122 `bamqc` after removal of marked duplicates: ( i ) depth of coverage as the # of unique reads mapping across all reference bases reported as a histogram and the mean # of unique reads mapping, ( ii ) breadth of coverage as % bases of reference genome covered by given #s of unique reads reported as a histogram and the % reference at 5× coverage by unique reads, and ( iii ) physical coverage or cumulative length of post- ﬁltering and metagenomic classi ﬁcation of collapsed reads expressed as a multiple of the expected 2.8B nucleotides composing the total canid genome size reported from ﬂow cytometry in coyotes, gray wolves, and domestic dogs 123,124 . As QualiMap does not exclude low mapping quality reads, unlike mosdepth, alignment statistics may be overestimated.

Sex inference of ancient specimens We found support for both DireGB and DireSP individuals as male after mapping libraries to the Greenland gray wolf reference using BWA ‘aln’ with stringent duplication removal (see Read Mapping 1B ). We calculated the Rx ratios 36,37 between chromosome X depth and autosomal mean depth of coverage and 95% conﬁdence interval (CI) lower and upper bounds for both individuals, and designated a 95% CI upper bound under 0.6 as male and a 95% CI lower bound over 0.8 as female 36,37 . The 95% CIs for the ratios of reads aligning to chromosome X versus autosomes placed between 0.539 to 0.554 for DireSP and 0.538 to 0.554 for DireGB.

While the ratios of reads mapping to the reference gray wolf X chromosome (1.0× for SP and 5.7× for GB) and autosomes (2× for SP and 10× for GB) supports these individuals as male, we did not observe similar coverage over the reference gray wolf Y chromosome (0.1× for SP and 0.4× for GB), which may be unsurprising: Y chromosomes are notoriously dicult to assemble given highly repetitive regions and high homology to the X chromosome, and is roughly 1/20th the size of the X chromosome, so mapping sequencing reads from ancient

17

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 18

DNA to the Y chromosome may be less stable than mapping to the X chromosome. Therefore, we took a closer look at reads mapping to Y-encoded genes as a secondary con ﬁrmation. We examined depth of coverage by dire wolf reads at the Y-encoded gene TSPY because it exhibits relatively high copy number variation among modern gray wolves 125 and has high copy numbers in both humans and cattle 126,127 . Thus, for males we expect to see a relatively higher abundance of reads aligning to TSPY than other windows of the assembled Y chromosome sequence. We identi ﬁed coordinates for the TSPY gene annotated on the Greenland gray wolf ( C. lupus ) assembly (GCA_905319855.2) at chrY:594247-596596 (ENSCAFG00865029519 in our gene annotation). We then calculated the depth of coverage by collapsed dire wolf reads aligned using bedtools 128 and mosdepth 121 across TSPY and 10,000 randomly sampled 2,349 nucleotide windows from assembled chromosome Y and 10,000 windows randomly sampled from any assembled autosome. We observed a high depth of coverage over TSPY (0.20× in DireSP and 0.72× in DireGB), relative to depths of coverage from equally-sized 2,349 nucleotide windows randomly sampled 10,000 times over the assembled gray wolf Y chromosome (0.103× ± SD: 0.397× in DireSP and 0.362× ± SD: 1.05× in DireGB), which, while lower than read depth over randomly sampled windows across autosomes (1.87× ± SD: 0.578× in DireSP and 10.6× ± SD: 2.86× in DireGB), supported the presence of male sex chromosomes.

Multi-canid pangenome construction and read mapping Ancient DNA mapped to multi-species pangenomes is expected to reduce reference bias from mapping to a single, divergent reference 89 . Thus we generated a multi-species pangenome of canids using the graph-based Minigraph-Cactus (MC) pipeline 129 , which starts from a reference assembly and iteratively introduces structural variation from aligned genome assemblies and, by excluding centromeric and highly repetitive sequences, can be more ecient for mapping sequencing read data 130 . We constructed the graph using the Greenland gray wolf (GenBank GCA_905319855.2) as our initial reference genome to be consistent with the coordinate system of our single reference alignments, and incorporated sequence variation from 6 chromosome-level assembled genomes that had the highest contig N50 and most complete set of chromosomes from two Canis spp. (domestic dog and dingo) and four other canid species and 3 scaold-level assemblies from dhole ( Cuon alpinus ), bush dog ( Speothos venaticus ), and red fox ( Vulpes vulpes ) for broad phylogenetic representation of the canid lineage (Table S2). We implemented a custom script (`buildPangenome.sh`) launched from the cactus docker image (v2.7.1) to construct a pangenome bidirectional sequence variation graph. We built a graph index and mapped unclassi ﬁed, trimmed reads using Girae 131 included in v g tools 132 . Only reads mapped to the mitochondria (MT) were extracted and utilized for dire wolf MT reconstruction. For downstream analysis, we surjected reads aligned to the multi-canid pangenome graph to the Greenland gray wolf (canLor) reference assembly. We observed that in pangenome alignments by Girae , short reads were occasionally aggressively clipped, resulting in a small fraction of bases within the overall read aligning. This phenomenon was most pronounced in the dire wolf individuals. We therefore ﬁltered pangenome alignments to remove reads which had more than 20% of their bases soft-clipped, and reads which had a non-clipped length of less than 35 bases.

18

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 19

Paleogenome reconstruction via iterative mapping and polishing We reconstructed the dire wolf nuclear genome by iteratively mapping and polishing to a divergent seed reference, the Greenland gray wolf (GenBank GCA_905319855.2), using ancient DNA to progressively assemble a less divergent reference sequence. We performed initial polishing from the DireGB BAM ﬁle (see Read Depth) by employing the Pilon polisher 133 in diploid mode to correct sequences for insertions, deletions, and single nucleotide polymorphisms (SNPs) and perform local reassembly of each chromosome or contig (`--changes --ﬁx all --diploid`).

For subsequent iterations, we allowed each library from DireGB to map to the polished sequences of autosomes, sex chromosomes, and unassembled contigs from the previous iteration, excluding the mitochondrial sequence which we assembled separately. We mapped unclassiﬁed and trimmed paired-end reads, merged and unmerged, using the BWA `mem` and standard deduplication with PALEOMIX (see Read Mapping 2A). Unmapped reads were discarded, libraries merged, and sequence polished. We carried out this process for 6 iterations in total and evaluated alignment statistics and quality metrics for each iteration (Table S4).

In order to reduce the computational load of assembling the mitochondrial genome and reduce reference bias in mapping, we ﬁrst mapped the complete libraries of DireGB to the canid Minigraph-Cactus pangenome to identify mitochondrial reads in our data. We surjected all reads that mapped to mitochondrial scaolds of the canid pangenome to the Greenland gray wolf mitochondrial (MT) sequence and reverse complemented the MT sequence to orient to strand with the D-loop control region. We then ﬁltered reads that were unmerged, unmapped, and not in primary alignment using samtools view -F 261 and removed duplicates with Picard `MarkDuplicates` function. The ﬁltered and duplicate-removed mitochondrial BAM was the ﬁnal input from which we extracted reads using samtools `fastq` function (-n), resulting in 29,104 reads from the DireGB libraries.

We used MIA 82 , an iterative mapping approach, to assemble the dire wolf mitochondrial genome with the gray wolf genome as the seed reference (NCBI Accession NC_008092.1), giving ﬂags (-c -C -k 14 -U) and -s with an ancient DNA substitution matrix. We ﬁltered the consensus assembly with “relaxed”, requiring 3× coverage and 66% consensus. All sites that did not pass these ﬁlters were set to ‘N’. For all downstream analyses, we used the relaxed ﬁltered assembly (Fig. S5). We estimated a 34.5× coverage mitochondrial consensus genome for DireSP and a 90.171× coverage mitochondrial genome for DireGB, with 3.35-21.26% N content (Table S3).

We performed a ﬁnal mapping of ancient DNA sequencing libraries to nuclear sequences resulting from the 6th iteration of mapping and polishing plus the iteratively assembled mitochondrial sequence. We aligned using BWA `aln` with standard PALEOMIX deduplication (Read Mapping 1A). We called pseudo-haploid consensus sequences by random read-sampling consensus and by variant-based consensus. In both cases, sequences below 5× unique read depth were rendered as hard-masked Ns, and we required a minimum base quality of BQ=30 and mapping quality of MQ=25. We performed random read-sampling using ANGSD 134 allele counting (`-doCounts 1 -minQ 30 -minMapQ 25 -dumpCounts 3`) followed by Consensify 135 to call consensus from reads mapped at 5× to 30× depth and 4/5 allelic support (`-min 5 -max 30 -n_matches 4 -n_random_reads 5`). We called a variant-based consensus to consider read depth and quality scores at variant and invariant sites. We called genotypes

19

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 20

using GATK 136 HaplotypeCaller emitting in gVCF mode, then converted to VCF using bcftools convert, ﬁltered variant calls (minimum QUAL=20, BQ=30, MQ=25, DP=5). In the case of overlapping alleles at variant sites, we enforced each genotype to represent homozygous genotype of the allele with highest depth, which should minimize bias from DNA damage-induced cytosine deamination producing false variant calls, using the remove-overlaps plugin for bcftools (https://github.com/samtools/bcftools/blob/develop/plugins/remove-overlaps.c). For variable sites, overlapping variants with the highest QUAL values were retained, and for overlapping sites that contain invariant sites, we generated pseudo-QUAL values using the highest QUAL value of a known variant site scaled by the depth of the invariant site prior to ﬁltering. Any sequences not covered in the resulting VCF were hard-masked as Ns in the ﬁnal consensus.

Gene annotation and orthology analysis We annotated genes using the fully automated FLAG tool (Troy et al. 2023) implemented in the `FLAG: Eukaryote Gene Annotation` workﬂow on the Form Bio platform (formbio.com). For the reconstructed dire wolf assembly, we included fully assembled transcript data from domestic dog (GCA_000002285.4), gray wolf (GCA_905319855.2), and dingo (GCA_003254725.2) as evidence for Liftover , Augustus , and Helixer . After employing a novel combine and ﬁlter algorithm, gene symbols were attributed to each model using Entap . We performed a BUSCO analysis using default settings and reported the total number of orthologs found in the mammalian database (n=9226). We assessed mean depth of coverage by all aligned short read libraries over all annotated genes (length-weighted average of mean depth per exon) in the reconstructed dire wolf genome sequence using mosdepth (Table S5).

Consensus calling from iterative reconstruction We mapped short read genomes of modern canid species sourced from publicly available data sets (Table S8), the new sequencing libraries from DireGB and DireSP individuals, and the previously published libraries of dire wolf individuals 12 to the reconstructed dire wolf genome reference (aenDir) using BWA `mem` with standard deduplication with ‘Mark Duplicates’ (Read Mapping 2A).

We called consensus sequences using a random read-sampling approach. We used ANGSD (`-doCounts 1 -dumpCounts 3`) to quantify depth of sequence variants with a minimum mapping and sequence quality of 25. We used these counts to generate a consensus genome using Consensify for the modern (`-min 10 -max 500 -n_matches 8 -n_random_reads 10`) and ancient (`-min 5 -max 30 -n_matches 4 -n_random_reads 5`) read sets. Note in both cases, a variable site was incorporated as a change to the reference only if 80% of alternative reads samples were in agreement. Such random sampling of ancient data also decreases the likelihood of incorporating false positive variants from DNA damage signals 138 .

Post hoc analysis of residual reference bias We assessed residual reference bias as a disproportionate rate of derived allele sharing with the seed reference Greenland gray wolf ( C. lupus orion aka “canLor”) genome (GenBank

20

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 21

GCA_905319855.2) 35 relative to the genome of a distinct gray wolf subspecies ( C. lupus aka “canLup”). We employed the ABBA/BABA D -statistic for introgression 38,84 to quantify derived allele sharing. We designated ancestral (A) alleles as those shared with outgroup species gray fox ( Urocyon cinereoargenteus aka “uroCin”) and/or red fox ( Vulpes vulpes aka “vulVul”), where for site patterns of derived (D) alleles considered concordant under the expected species tree, `(((canLor,canLup), aenDir), outgroup/vulVul/uroCin)` would be `(((D,D),A),A)` or `(((D,D),D),A)` and discordant as `(((D,A),D),A)` and `(((A,D),D),A)`. We assessed these site patterns using haploid calls from long sequence alignment over syntenic regions to reference genomes of outgroup canid species varying in divergence.

First, we performed a long assembly-to-reference mapping with minimap2 139 using the preset for up to 10% sequence divergence (`asm10`), and applied the minimap2 `paftools call` function to identify aligned regions (R lines) and sequence variants (V lines) relative to outgroup assemblies from most-to-least diverged from or within the wolf-like clade: the gray fox reference assembly (GenBank GCA_032313775.1) 112 and the red fox reference assembly (GenBank GCA_964106925.2) 111 , the maned wolf (GenBank GCA_028533335.1, Chrysocyon brachyurus aka “chrBra”) reference assembly 107 , and the African wild dog ( Lycaon pictus aka “lycPic”) reference assembly (Yggdrasil DB12) 114 . To those reference genomes, we aligned assembled genome sequences of outgroups red fox and gray fox, when not used as reference; the seed-reference Greenland gray wolf ( C. lupus orion aka canLor) and a non-reference genome assembled for a North American wolf (GenBank GCA_003160815.1, C. lupus aka canLup) 113 ; and the pseudo-haploid consensus genome sequences of dire wolf individual DireGB called from random read-sampling ANGSD + Consensify from alignment to either the reconstructed ancient dire wolf ( Aenocyon dirus aka “aenDir”) genome reference (“iterative”) or its seed-reference sequence of Greenland gray wolf (“divergent”).

We restricted analyses to regions over each outgroup genome aligned by gray wolf and dire wolf genomes, comprising a total of 171,599,977 gray fox aligned sequence bases, 247,756,445 maned wolf aligned bases, and 244,509,933 African wild dog aligned bases, where these taxa should have long-range synteny for sequence alignment and variant calling. We then computed global D -statistic as (# ADDA sites - # DADA sites)/(# ADDA sites + # DADA sites) over canid syntenic regions of respective reference assemblies. We computed standard errors via a block jackknife procedure 38 to divide the outgroup genomes into 1 megabase equally-spaced contiguous blocks, and adjusted global D -statistics by this standard error to obtain Z-scores (Table S6).

We mapped DireSP sequence reads using BWA mem as described above to either the iteratively reconstructed dire wolf or gray wolf reference assemblies. Using the Rsamtools package (v2.22.0), we sampled 500,000 reads from each alignment ﬁle and parsed their cigar strings to calculate the identity-by-state (IBS) for each read. IBS was de ﬁned as the (number of aligned bases (M) - number of mismatches (NM))/aligned sequence length (M). Reads were binned by their aligned sequence lengths and average IBS was calculated for each bin.

Mitochondrial and nuclear phylogenetic analysis In order to place our newly assembled mitochondrial sequence in phylogenetic context, we downloaded publicly available canid mitochondrial genomes (Table S9). We used rotate 140 to ensure all circularized sequences started at the same position and aligned the sequences with MUSCLE v3.8.1 141 . We used RAxML v8.2.12 142 to build a maximum likelihood phylogeny (-m

21

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 22

GTRGamma -f a -# 100 -p 12345 -x 12345) under the GTRGAMMA substitution model with 100 bootstrap replicates. The dire wolf mitochondrial genomes formed a clade outside of other wolf-like canids (Fig. S5) in a topology similar to that estimated previously 12,20,143 . To localize the placement of the dire wolf in the broader canid phylogeny, we set out to build a nuclear phylogeny using neutral windows of genomic sequence. We de ﬁned neutral windows using bedtools makewindows (-w 25000 -s 1025000 -i winnum), which deﬁned uniquely named 25 Kb loci at least 1 Mb apart. Windows containing any annotated protein coding genes were removed using bedtools intersect (-v). This resulted in 1,332 possible loci to build trees from.

To assess canid sequence divergence over these windows, we aligned 14 representative canid data sets (Table S8), along with DireGB and DireSP, against the iteratively assembled paleogenome reference and obtained consensus sequencing using the ANGSD + Consensify approach outlined above. Sequences over the neutral windows were extracted and N content was measured using the “letterFrequency” function from the BioStrings package in R. Only windows that exhibited <= 50% missingness across all species (n=924) were retained for further analysis. Sequences that passed these thresholds were converted to multiple sequence alignments using MAFFT and trees were built from each loci using IQTree2 with 100 bootstraps each with a GTR + I + G substitution model 144,145 . Each loci tree was merged and provided as input for ASTRAL II 146,147 to call a ﬁnal consensus species tree.

Divergence time calibration We used MCMCtree v4.10.7 from the PAML package 148 to perform divergence time estimation. We used an autocorrelated-rates relaxed clock model with approximate likelihood calculation and the HKY85 149 substitution model with gamma-distributed (ﬁve categories) rate heterogeneity among sites. We used gamma priors for kappa (shape=2, scale=10) and alpha (1, 10) in the substitution model, and gamma-Dirichlet priors for the overall substitution rate (shape=2, scale=40, concentration=1) and rate drift parameter (2, 20, 1). These priors were selected to provide diuse yet biologically realistic constraints on the nucleotide transition/transversion ratio, among-site rate variation, overall rate of molecular evolution, and rate drift (allowing for changes in evolutionary rate from parent to ospring branches) in the autocorrelated relaxed clock model, respectively.

We used two nodes as calibration points with priors based on previous studies: the root age calibrated for gray foxes ( Urocyon spp.) (same root calibration as Perri et al. 2021) versus other canids 12 and the South American canid node represented by bush dog and maned wolf 107 . Lower and upper bounds for the root age were 10.3 Mya ( Metalopex macconnelli ) and 20 Mya ( Leptocyon spp.), and 1.35-4.34 Mya for the South American canid node. For both nodes, we used log-normal distributions as priors, with 0.025 for location and scale parameters. This approximated a uniform distribution between upper and lower bounds, while softening the bounds with low-probability tails (see prior distributions of node ages in Fig. S9). The MCMC analysis was run for 1,000,000 iterations, with the ﬁrst 10,000 discarded as burn-in, and samples collected every 100 iterations, resulting in 10,000 samples. We created visualizations of the resulting tree with divergence time estimates using the MCMCtreeR package 150 in R, showing both node age uncertainty and posterior distributions of node ages (Fig. 3, Table S10).

22

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 23

Sequence variant calling We detected germline variants against the reconstructed dire wolf reference (aenDir) from BAMs of 35 modern canids (Table S8) and DireGB and DireSP using a joint-calling work ﬂow implemented by GATK4 136 : HaplotypeCaller for local re-assembly of haplotypes, GenomicsDBImport for gVCF consolidation, and GenotypeGVCFs for joint variant calling. We also generated a variant call set comprising transversions only by ﬁltering with BCFtools option, `-i '((REF="A" & ALT~"[TC]") | (REF="T" & ALT~"[AG]") | (REF="C" & ALT~"[AG]") | (REF="G" & ALT~"[TC]")) & (strlen(REF)=1 & strlen(ALT)=1)'`, which totaled 41,774,524 SNPs. We also called variants from the multi-canid pangenome alignments surjected to the Greenland gray wolf reference (canLor) using the GPU-optimized NVIDIA® Clara™ Parabricks implementation of DeepVariant 151 for 23 modern canids (Table S8) and both dire wolves. DeepVariant gVCFs were combined using GLNexus 152 with the ‘DeepVariant_unﬁltered’ conﬁguration to avoid biasing allele frequencies of non-gray wolf species. We then set as missing any genotype with GQ<10 using the bcftools setGT plugin and ﬁltered any site where fewer than 10 individuals were genotyped.

Sequence distance, diversity, and divergence We estimated the sequence distance of dire wolves from extant canids using the formula: 2 x (mutations/generation) x (divergence time / generation time), assuming a mutation rate of 4 billion substitutions per generation calibrated for Canis species 41 , a generation time between 5 and 8 years for black-backed jackals and wolf-like canids, respectively 42 , and a divergence time of 5.8 Mya 12 . We assessed sequence similarity in two ways. First, we ﬁltered the ﬁnal species variant call set (described above) for homozygous transversions by species using BCFTools query and compared each using a custom script. Each species in a comparison had to have the same site vary in the same way to be considered a shared allele. We then subtracted the total dierences from the non-N sequence length of the reconstructed dire wolf (2,310,286,013 nucleotides) and expressed a proportion of total available sequence. In parallel, we estimated sequence distance using Mash 43 . For each consensus sequence, we used mash sketch to hash the genome into 32-mers and assessed common 32-mers in each species in a pairwise manner.

Gene flow and admixture graph modeling We converted autosomal variants to EIGENSTRAT format using vcf2eigenstrat from SequenceTools (https://github.com/stschi/sequenceTools) for both the GATK4 jointly-called variants against the dire wolf consensus genome and DeepVariant variants called against the gray wolf genome from the surjected pangenome alignments. To account for dierential coverage across individuals, we then made pseudohaploid genotype calls for each individual for the full set of variants using a read sampling procedure modeled after Consensify 135 that controls for biases arising from dierential sequencing coverage and mapping errors (Fig. S10, S19) by randomly selecting 3 reads at each site, or two reads at positions covered by exactly two reads, and requiring at least two bases among the selected reads match. Sites with fewer than two reads were set to missing. This procedure resulted in 33.91M aenDir-mapped and 35.06M pangenome-mapped autosomal, biallelic, mappabilityﬁltered

23

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 24

transversions for 17 populations represented by 37 individuals. Ultimately, alignment to the reconstructed dire wolf reference increased reference bias (attraction toward DireGB) in a manner that pangenome alignment did not (Fig. S12, S19), particularly for more divergent outgroups. As admixture tests can be particularly sensitive to such biases 25 , we focused pangenome-derived results for describing canid population history and gene ﬂow, though note that patterns and ratios of admixture in the dire wolf lineage obtained from alignments to the reconstructed dire wolf genome were similar to those inferred from pangenome aligned reads (Fig. S13).

We used Dstatistics to assess allele sharing between dire wolves and extant Canini lineages, using either red foxes or gray foxes as an outgroup D -statistics were calculated using admixtools2 v2.0.6 44 . We then used the ﬁndgraphs function from admixtools2 to search for possible well-ﬁtting admixture graph models across varying subsets of canid species. For each set of species, we performed 100 separate iterations of ﬁndgraphs across 0-8 admixture events, and extracted the graph with the lowest log-likelihood score as the bestﬁtting model across each iteration. In cases where less complex graphs provided adequate ﬁts for a given set of species (worst residual of Z< 3), we did not model additional admixture events.

We began with a minimal set of species representing major canid lineages: dire wolves (n=2), Eurasian gray wolves (n=2), coyotes (n=4), dholes (n=2), African wild dogs (n=3), maned wolves (n=2), and red foxes (n=2) as an outgroup. We also explored similar models but with either gray foxes or both red and gray foxes as outgroups. We also explored graphs in which dire wolves were each modeled individually, rather than grouped into a population. Beyond this minimal species set, we also modeled a complex graph which featured African jackals (side-striped jackal and black-backed jackal), along with an additional South American canid lineage (Pampas fox), and a North American gray wolf population (Yellowstone). Finally, we assessed admixture graphs which featured only extant canid lineages and included populations with known history of admixture (golden jackals and red wolves), to assess the sensitivity of this method to detect known instances of admixture among canid lineages. Diversifying positive selection analysis We extracted all protein-coding nucleotide and translated amino acid sequences using gread 153 on the pseudo-haploid consensus genome sequences of modern canid species and ancient dire wolf aligned to reconstructed dire wolf paleogenome and called from the random read-sampling ANGSD + Consensify method, which maintained identical genomic coordinates for gene annotation. We selected the longest coding transcript per distinct gene ID with <20% missing amino acids when translated using seqkit 154 , yielding 13,830 protein-coding genes for positive selection tests. We ran HyPhy analysis programs BUSTED and FitMG94 in local and global modes on aligned CDS and the consensus species tree with dire wolf (DireGB) designated as foreground branch:

(uroCin_SAMN04495241,(vulVul_marble_Fox4,((chrBra_SAMN27257233,speVen_SAMN272 57238)1:3.795527799956191,(DireGB,(lupMes_SAMEA5411047,(lycPic_SAMN09924608,(cuoAl p_SAMN10180424,(canRuf_SAMN02921317,(canLup_SAMN02921310,canBai_SAMN0292131 4)1:0.11623018789473029)1:4.188188635818363)1:0.39859685674669815)1:0.323225231778891 )1:0.1826891919310381)1:1.4874594855476162)1:6.424328629404271):0.0);

24

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 25

We tested protein-coding genes using the BUSTED model for episodic positive selection to identify where at least one site experienced positive selection in the dire wolf branch. Additionally, we extracted from the FitMG94 output the branch-speci ﬁc dN/dS rate estimates under a free-ratio model (`--type local`) and respective non-synonymous and synonymous branch lengths for dire wolf and 10 other canids and the likelihood-ratio test for overall dN/dS != 1 across the entire species tree using a one-ratio model (`--type global`). We considered those genes with tree-wide, one-ratio dN/dS estimate > 2 as indicating pervasive positive selection over the canid lineage.

We ﬂagged genes as under episodic positive aka diversifying selection in the dire wolf branch that exhibited p < 0.05 in the BUSTED test and which ﬂagged at least a single codon of evidence ratio > 2, and excluding genes for which the dire wolf consensus codon did not disagree with dire wolf reconstructed reference codon and neither the focal dire wolf codon nor root canid codon contained missing bases (Ns). The process of consensus calling at sites of low depth (<5×) or heterozygous alleles introduces missing bases that could result in misassignment of codons for the focal branch (e.g., dire wolf) and/or root branch (ancestral canid), which would in turn lead to false positives as HyPhy BUSTED needs only a single site to reject ﬁtting the constrained model (no positive selection). Furthermore, BUSTED is insensitive to synonymous substitution rate variation (a constant dS=1 is held in all cases).

We also ﬂagged additional genes as having strong evidence for positive aka diversifying selection in the dire wolf branch, if these had free-ratio maximum likelihood estimates of dN/dS > 2 from the standard MG94 codon evolution model 155 implemented in the HyPhy FitMG94 program. We ﬁltered out such genes depleted of synonymous substitutions where dS > 0.001, to avoid extreme estimates, and no excess saturation of substitutions where dS or dN was greater than 4.0, and which exhibited overall dN/dS estimated under the one-ratio model meeting statistical signiﬁcance (α=0.05) in the log-ratio test given degrees of freedom=(estimated parameters in the standard MG94 model - estimated parameters in the nucleotide GTR model) > 0 under the chi-squared distribution. We excluded those that exhibited extremely saturated rates of substitutions (dS or dN > 4.0) or depleted rates of synonymous substitutions (dS < 0.001), because dN/dS ratios could be inappropriately estimated or result from misalignment of reads 156 . We report all protein-coding genes tested and respective statistics from HyPhy BUSTED and FitMG94 programs in Table S11. As BUSTED evidence ratios only indicate whether a site better ﬁts under the alternative, unconstrained model that permits positive selection, rather than providing direct evidence site-speci ﬁc positive selection, we also ran HyPhy FEL ( ﬁxed eects indicating pervasive selection) and MEME (mixed eects aka partitioned selection) maximum likelihood models over all 80 genes under diversifying positive selection in dire wolves.

We performed gene set enrichment analysis using g:Proﬁler (`gproﬁler2`) 157 and the domestic dog ortholog reference set (organism: “clfamiliaris”), testing the combined suite of 80 genes under episodic diversifying selection against the background set of 5,497 genes tested that pass all aforementioned ﬁlters. Corrections for multiple testing were performed using the Benjamini-Hochberg procedure.

We contextualized dire wolf speciﬁc substitutions using InterProScan to predict protein functional domains 53 and ESM1b model to predict substitution impact via in silico mutagenesis 54 over each dire wolf protein sequence. We replaced all invalid amino acid residues (`X` introduced from N content in coding sequences) with a neutral amino acid,

25

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 26

glycine (`G`), to enable modeling. For protein sequences longer than 1,022 amino acids that hit the sequence limit of the ESM1b model, we performed tiled predictions as in Brandes et al. 2023 54 to obtain weighted average eect scores. We examined residues harboring dire wolf speciﬁc amino acids and compared to the canid-majority amino acid at each residue. As the “wild type” sequence modeled is sourced from the reconstructed dire wolf reference (aenDir), we considered the positive or negative pseudo log-likelihood scores between the consensus amino acid assigned to the dire wolf branch in BUSTED and the canid majority amino acid, depending which amino acid matched the root in BUSTED analysis. We report domain and eect predictions for ﬂagged residues modeled in Table S12.

DATA AVAILABILITY The sequencing data underlying this study are available in the NCBI Sequence Read Archive, and can be accessed with BioProject accession PRJNA1222369. For the HyPhy analyses, we included a tarball of the multiple sequence alignment FASTAs of the coding sequences for all 80 genes under diversifying selection in the dire wolf branch as well as the HyPhy results JSON ﬁles from BUSTED, FitMG94, FEL, and MEME programs. These data and all supplemental tables are available in a Zenodo repository associated with this study under the BioRxiv DOI.

ACKNOWLEDGEMENTS We thank the Center for Genome Innovation and the Computational Biology Core in the Institute for Systems Genomics at the University of Connecticut for sequencing and data management support. We thank the Bureau of Reclamation for sampling permissions of the Idaho individual that produced DireGB (IMNH 48001/52) and sta at the Idaho Museum of Natural History for supporting resampling of specimens under their care. We thank the sta including Glenn Storrs and Cameron Schwalbach of the Cincinnati Museum Center for additional sampling of the DireSP specimen (VP1737). We also thank colleagues who lent their time and expertise to provide valuable comments on this work.

AUTHOR CONTRIBUTIONS Conceptualization: G.G., K.M.P, J.O., M.CJ., B.L.C, B.S. Data curation: G.G., K.M.P, J.O., B.L.C. Formal analysis: G.G., K.M.P, J.O., C.H., M.CJ., W.T. Funding acquisition: B.L. Investigation: S.B., C.C., J.M., B.R.P, P.G.S.G., S.J.H., S.S., W.S., R.J.O, W.T. Methodology: G.G., K.M.P, J.O., M.CJ., B.L.C, B.S. Project administration: G.G., K.M.P Resources: N.A, S.B., B.R.P., J.M., S.J.H., R.J.O, S.S., W.S., C.C., L.K., T.M., B.S. Software: N.A., G.G., K.M.P, J.O., M.CJ., Supervision: B.L.C, B.S.

26

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 27

Validation: G.G., K.M.P., J.O. Visualization: G.G., K.M.P, J.O., M.CJ, C.H. Writing – original draft: K.M.P., G.G. Writing – review & editing: G.G., K.M.P., J.O. M.CJ., C.C., N.A, W.T, O.F, S.B., T.M., L.K., J.K., A.D., M.C., J.B.P, S.O., R.J.O, S.J.H., P.G.S.G., S.S., W.S., J.M., B.R.P, G.L., L.F, A.P., A.T.L, M.S.S., M.T.P.G, S.G., B.vH., E.K.K., C.E.M., L.D, B.L.C., G.R.R.M., G.C, B.S.

CONFLICTS OF INTEREST Authors aliated with Colossal Biosciences and/or Form Bio may hold stock and/or stock options in these companies.

E.K.K., C.E.M, B.vH., R.J.O., M.T.P.G, and L.D. are members of the scientiﬁc advisory board at Colossal Biosciences.

M.S.S. is a consultant of Colossal Biosciences.

G.R.R.M is an Investor and Cultural Advisor for Colossal Biosciences.

G.C. is co-founder of Colossal Biosciences and others available at arep.med.harvard.edu/ t

SUPPLEMENTAL INFORMATION Document S1: Figures S1-19 Table S1: Sequencing library statistics for all ancient samples. Table S2: Reference assembly accessions used in study Table S3: Alignment statistics for all ancient samples Table S4: Iterative mapping and polishing statistics for paleogenome reconstruction Table S5: Gene annotation information for paleogenome reconstruction Table S6: Allele sharing statistics for residual reference bias assessment Table S7: Genetic distance metrics Table S8: Short read dataset accessions used in study Table S9: Mitochondrial genome accessions used in study Table S10: Age estimates for nodes in phylogenetic tree Table S11: Summary of admixture methods and models Table S12: Selection test results by gene Table S13: Genes with dire wolf speciﬁc substitutions

27

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 28

REFERENCES

1. Mead, J.I., Manganaro, C., Repenning, C.A., and Agenbroad, L.D. (1996). Early

Rancholabrean mammals from Salamander Cave, Black Hills, South Dakota. In Palaeoecology and Palaeoenvironments of Late Cenozoic Mammals (University of Toronto Press), pp. 458–482. https://doi.org/10.3138/9781487574154-023.

2. O’Keefe, F.R., Dunn, R.E., Weitzel, E.M., Waters, M.R., Martinez, L.N., Binder, W.J., Southon,

J.R., Cohen, J.E., Meachen, J.A., DeSantis, L.R.G., et al. (2023). Pre–Younger Dryas megafaunal extirpation at Rancho La Brea linked to ﬁre-driven state shift. Science 381 , eabo3594. https://doi.org/10.1126/science.abo3594.

3. Anyonge, W., and Baker, A. (2006). Craniofacial morphology and feeding behavior in

Canis dirus , the extinct Pleistocene dire wolf. J. Zool. 269 , 309–316. https://doi.org/10.1111/j.1469-7998.2006.00043.x.

4. De Latorre, D.V., and Marshall, C.R. (2024). Evolutionary allometry of the canid baculum

(Carnivora: Mammalia). Biol. J. Linn. Soc., blae048. https://doi.org/10.1093/biolinnean/blae048.

5. Patterson, B.D., and Thaeler, C.S. (1982). The Mammalian Baculum: Hypotheses on the

Nature of Bacular Variability. J. Mammal. 63 , 1–15. https://doi.org/10.2307/1380665.

6. O’Keefe, F.R., Meachen, J., Fet, E.V., and Brannick, A. (2013). Ecological determinants of

clinal morphological variation in the cranium of the North American gray wolf. J. Mammal. 94 , 1223–1236. https://doi.org/10.1644/13-MAMM-A-069.

7. Dundas, R.G. (1999). Quaternary records of the dire wolf, Canis dirus , in North and South America. Boreas 28 , 375–385. https://doi.org/10.1111/j.1502-3885.1999.tb00227.x.

8. Anyonge, W. and Roman, Chris (2006). New Body Mass Estimates for Canis Dirus, the

Extinct Pleistocene Dire Wolf. J. Vertebr. Paleontol. 26 , 209–212. https://doi.org/10.1671/0272-4634(2006)26[209:NBMEFC]2.0.CO;2.

9. Harris, A.H. (1981). Kurten, B., and E. Anderson. PLEISTOCENE MAMMALS OF NORTH

AMERICA. Columbia Univ. Press, New York, 442 pp., 1980. Price $42.50. J. Mammal. 62 , 653–654. https://doi.org/10.2307/1380422.

10. Tedford, R.H., Wang, X., and Taylor, B.E. (2009). Phylogenetic Systematics of the North

American Fossil Caninae (Carnivora: Canidae). Bull. Am. Mus. Nat. Hist. 325 , 1–218. https://doi.org/10.1206/574.1.

11. Nowak, R.M. (2010). Wolf Evolution and Taxonomy. In Wolves: Behavior, Ecology, and

Conservation, L. D. Mech and L. Boitani, eds. (University of Chicago Press).

12. Perri, A.R., Mitchell, K.J., Mouton, A., Álvarez-Carretero, S., Hulme-Beaman, A., Haile, J.,

Jamieson, A., Meachen, J., Lin, A.T., Schubert, B.W., et al. (2021). Dire wolves were the last

28

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 29

of an ancient New World canid lineage. Nature 591 , 87–91. https://doi.org/10.1038/s41586-020-03082-x.

13. Sotnikova, M., and Rook, L. (2010). Dispersal of the Canini (Mammalia, Canidae: Caninae)

across Eurasia during the Late Miocene to Early Pleistocene. Quat. Int. 212 , 86–97. https://doi.org/10.1016/j.quaint.2009.06.008.

14. Meachen, J.A., and Samuels, J.X. (2012). Evolution in coyotes ( Canis latrans ) in response to the megafaunal extinctions. Proc. Natl. Acad. Sci. 109 , 4191–4196. https://doi.org/10.1073/pnas.1113788109.

15. Meachen, J.A., Janowicz, A.C., Avery, J.E., and Sadleir, R.W. (2014). Ecological Changes in

Coyotes (Canis latrans) in Response to the Ice Age Megafaunal Extinctions. PLoS ONE 9 , e116041. https://doi.org/10.1371/journal.pone.0116041.

16. Nowak, R.M. (1979). North American Quaternary Canis (Museum of Natural History,

University of Kansas) https://doi.org/10.5962/bhl.title.4072.

17. Heintzman, P.D., Froese, D., Ives, J.W., Soares, A.E.R., Zazula, G.D., Letts, B., Andrews, T.D.,

Driver, J.C., Hall, E., Hare, P.G., et al. (2016). Bison phylogeography constrains dispersal and viability of the Ice Free Corridor in western Canada. Proc. Natl. Acad. Sci. 113 , 8057–8063. https://doi.org/10.1073/pnas.1601077113.

18. Koblmüller, S., Vilà, C., Lorente‐Galdos, B., Dabad, M., Ramirez, O., Marques‐Bonet, T.,

Wayne, R.K., and Leonard, J.A. (2016). Whole mitochondrial genomes illuminate ancient intercontinental dispersals of grey wolves ( Canis lupus ). J. Biogeogr. 43 , 1728–1738. https://doi.org/10.1111/jbi.12765.

19. Loog, L., Thalmann, O., Sinding, M.S., Schuenemann, V.J., Perri, A., Germonpré, M.,

Bocherens, H., Witt, K.E., Samaniego Castruita, J.A., Velasco, M.S., et al. (2020). Ancient DNA suggests modern wolves trace their origin to a Late Pleistocene expansion from Beringia. Mol. Ecol. 29 , 1596–1610. https://doi.org/10.1111/mec.15329.

20. Koepﬂi, K.-P., Pollinger, J., Godinho, R., Robinson, J., Lea, A., Hendricks, S., Schweizer,

R.M., Thalmann, O., Silva, P., Fan, Z., et al. (2015). Genome-wide Evidence Reveals that African and Eurasian Golden Jackals Are Distinct Species. Curr. Biol. 25 , 2158–2165. https://doi.org/10.1016/j.cub.2015.06.060.

21. vonHoldt, B.M., Cahill, J.A., Fan, Z., Gronau, I., Robinson, J., Pollinger, J.P., Shapiro, B.,

Wall, J., and Wayne, R.K. (2016). Whole-genome sequence analysis shows that two endemic species of North American wolf are admixtures of the coyote and gray wolf. Sci. Adv. 2 , e1501714. https://doi.org/10.1126/sciadv.1501714.

22. Gopalakrishnan, S., Sinding, M.-H.S., Ramos-Madrigal, J., Niemann, J., Samaniego

Castruita, J.A., Vieira, F.G., Carøe, C., Montero, M.D.M., Kuderna, L., Serres, A., et al. (2018). Interspeciﬁc gene ﬂow shaped the evolution of the genus Canis. Curr. Biol. 28 , 3441-3449.e5. https://doi.org/10.1016/j.cub.2018.08.041.

29

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 30

23. Westbury, M.V., Hartmann, S., Barlow, A., Preick, M., Ridush, B., Nagel, D., Rathgeber, T.,

Ziegler, R., Baryshnikov, G., Sheng, G., et al. (2020). Hyena paleogenomes reveal a complex evolutionary history of cross-continental gene ﬂow between spotted and cave hyena. Sci. Adv. 6 , eaay0456. https://doi.org/10.1126/sciadv.aay0456.

24. Westbury, M.V., Barnett, R., Sandoval-Velasco, M., Gower, G., Vieira, F.G., de Manuel, M.,

Hansen, A.J., Yamaguchi, N., Werdelin, L., Marques-Bonet, T., et al. (2021). A genomic exploration of the early evolution of extant cats and their sabre-toothed relatives. Open Res. Eur. 1 , 25. https://doi.org/10.12688/openreseurope.13104.2.

25. Günther, T., and Nettelblad, C. (2019). The presence and impact of reference bias on

population genomic studies of prehistoric human populations. PLoS Genet. 15 , e1008302. https://doi.org/10.1371/journal.pgen.1008302.

26. Prüfer, K., Stenzel, U., Hofreiter, M., Pääbo, S., Kelso, J., and Green, R.E. (2010).

Computational challenges in the analysis of ancient DNA. Genome Biol. 11 , R47. https://doi.org/10.1186/gb-2010-11-5-r47.

27. Liu, Y., Wang, M., Chen, P., Wang, Z., Liu, J., Yao, L., Wang, F., Tang, R., Zou, X., and He, G.

(2021). Combined Low-/High-Density Modern and Ancient Genome-Wide Data Document Genomic Admixture History of High-Altitude East Asians. Front. Genet. 12 , 582357. https://doi.org/10.3389/fgene.2021.582357.

28. Shapiro, B., and Hofreiter, M. (2014). A paleogenomic perspective on evolution and gene

function: new insights from ancient dna. Science 343 , 1236573. https://doi.org/10.1126/science.1236573.

29. Akopyan, M., Genchev, M., Armstrong, E.E., and Mooney, J.A. (2024). Divergent reference

genomes compromise the reconstruction of demographic histories, selection scans, and population genetic summary statistics. Preprint at bioRxiv, https://doi.org/10.1101/2024.11.26.625554 https://doi.org/10.1101/2024.11.26.625554.

30. O’Keefe, F., Binder, W., Frost, S., Sadleir, R., and Van Valkenburgh, B. (2014). Cranial

morphometrics of the dire wolf, Canis dirus , at Rancho La Brea: temporal variability and its links to nutrient stress and climate. Palaeontol. Electron. https://doi.org/10.26879/437.

31. Schmökel, H., Farrell, A., and Balisi, M.F. (2023). Subchondral defects resembling

osteochondrosis dissecans in joint surfaces of the extinct saber-toothed cat Smilodon fatalis and dire wolf Aenocyon dirus. PloS One 18 , e0287656. https://doi.org/10.1371/journal.pone.0287656.

32. O’Keefe, F., Binder, W., Frost, S., Sadleir, R., and Van Valkenburgh, B. (2014). Cranial

morphometrics of the dire wolf, Canis dirus , at Rancho La Brea: temporal variability and its links to nutrient stress and climate. Palaeontol. Electron. https://doi.org/10.26879/437.

33. Tankersley, K.B. (1997). Sheriden: A Clovis cave site in eastern North America.

Geoarchaeology 12 , 713–724.

30

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 31

https://doi.org/10.1002/(SICI)1520-6548(199709)12:6<713::AID-GEA9>3.0.CO;2-1.

34. Scott, W.E., Pierce, K.L., Bradbury, J.P., and Forester, R.M. (1982). Revised Quaternary

stratigraphy and chronology in the American Falls area, southeastern Idaho. Cenozoic Geol. Ida. Ida. Bur. Mines Geol. Bull. 26 , 581–595.

35. Sinding, M.-H.S., Gopalakrishnan, S., Raundrup, K., Dalén, L., Threlfall, J., Darwin Tree of

Life Barcoding collective, Wellcome Sanger Institute Tree of Life programme, Wellcome Sanger Institute Scientiﬁc Operations: DNA Pipelines collective, Tree of Life Core Informatics collective, Darwin Tree of Life Consortium, et al. (2021). The genome sequence of the grey wolf, Canis lupus Linnaeus 1758. Wellcome Open Res. 6 , 310. https://doi.org/10.12688/wellcomeopenres.17332.1.

36. Mittnik, A., Wang, C.-C., Svoboda, J., and Krause, J. (2016). A Molecular Approach to the

Sexing of the Triple Burial at the Upper Paleolithic Site of Dolní Věstonice. PLOS ONE 11 , e0163019. https://doi.org/10.1371/journal.pone.0163019.

37. De Flamingh, A., Coutu, A., Roca, A.L., and Malhi, R.S. (2020). Accurate sex identiﬁcation

of ancient elephant and other animal remains using low-coverage DNA shotgun sequencing data. G3 GenesGenomesGenetics 10 , 1427–1432. https://doi.org/10.1534/g3.119.400833.

38. Meyer, M., Kircher, M., Gansauge, M.-T., Li, H., Racimo, F., Mallick, S., Schraiber, J.G., Jay,

F., Prüfer, K., De Filippo, C., et al. (2012). A High-Coverage Genome Sequence from an Archaic Denisovan Individual. Science 338 , 222–226. https://doi.org/10.1126/science.1224344.

39. Hahn, C., Bachmann, L., and Chevreux, B. (2013). Reconstructing mitochondrial genomes

directly from genomic next-generation sequencing reads—a baiting and iterative mapping approach. Nucleic Acids Res. 41 , e129. https://doi.org/10.1093/nar/gkt371.

40. Dolenz, S., Van Der Valk, T., Jin, C., Oppenheimer, J., Sharif, M.B., Orlando, L., Shapiro, B.,

Dalén, L., and Heintzman, P.D. (2024). Unravelling reference bias in ancient DNA datasets. Bioinformatics 40 , btae436. https://doi.org/10.1093/bioinformatics/btae436.

41. Skoglund, P., Ersmark, E., Palkopoulou, E., and Dalén, L. (2015). Ancient wolf genome

reveals an early divergence of domestic dog ancestors and Aamixture into high-latitude Breeds. Curr. Biol. 25 , 1515–1519. https://doi.org/10.1016/j.cub.2015.04.019.

42. Paciﬁci, M., Santini, L., Di Marco, M., Baisero, D., Francucci, L., Grottolo Marasini, G.,

Visconti, P., and Rondinini, C. (2013). Generation length for mammals. Nat. Conserv. 5 , 89–94. https://doi.org/10.3897/natureconservation.5.5734.

43. Ondov, B.D., Treangen, T.J., Melsted, P., Mallonee, A.B., Bergman, N.H., Koren, S., and

Phillippy, A.M. (2016). Mash: fast genome and metagenome distance estimation using MinHash. Genome Biol. 17 , 132. https://doi.org/10.1186/s13059-016-0997-x.

31

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 32

44. Maier, R., Flegontov, P., Flegontova, O., Işıldak, U., Changmai, P., and Reich, D. (2023). On

the limits of ﬁtting complex models of population history to f-statistics. eLife 12 , e85492. https://doi.org/10.7554/eLife.85492.

45. Gopalakrishnan, S., Sinding, M.-H.S., Ramos-Madrigal, J., Niemann, J., Samaniego

Castruita, J.A., Vieira, F.G., Carøe, C., Montero, M. de M., Kuderna, L., Serres, A., et al. (2018). Interspeciﬁc Gene Flow Shaped the Evolution of the Genus Canis . Curr. Biol. 28 , 3441-3449.e5. https://doi.org/10.1016/j.cub.2018.08.041.

46. Chavez, D.E., Gronau, I., Hains, T., Kliver, S., Koepﬂi, K.-P., and Wayne, R.K. (2019).

Comparative genomics provides new insights into the remarkable adaptations of the African wild dog (Lycaon pictus). Sci. Rep. 9 , 8329. https://doi.org/10.1038/s41598-019-44772-5.

47. Heppenheimer, E., Brzeski, K.E., Hinton, J.W., Chamberlain, M.J., Robinson, J., Wayne, R.K.,

and vonHoldt, B.M. (2020). A genome-wide perspective on the persistence of red wolf ancestry in southeastern canids. J. Hered. 111 , 277–286. https://doi.org/10.1093/jhered/esaa006.

48. Vilaça, S.T., Donaldson, M.E., Benazzo, A., Wheeldon, T.J., Vizzari, M.T., Bertorelle, G.,

Patterson, B.R., and Kyle, C.J. (2023). Tracing Eastern Wolf Origins From Whole-Genome Data in Context of Extensive Hybridization. Mol. Biol. Evol. 40 , msad055. https://doi.org/10.1093/molbev/msad055.

49. Koepﬂi, K.-P., Pollinger, J., Godinho, R., Robinson, J., Lea, A., Hendricks, S., Schweizer,

R.M., Thalmann, O., Silva, P., Fan, Z., et al. (2015). Genome-wide Evidence Reveals that African and Eurasian Golden Jackals Are Distinct Species. Curr. Biol. 25 , 2158–2165. https://doi.org/10.1016/j.cub.2015.06.060.

50. Murrell, B., Weaver, S., Smith, M.D., Wertheim, J.O., Murrell, S., Aylward, A., Eren, K., Pollner,

T., Martin, D.P., Smith, D.M., et al. (2015). Gene-wide identiﬁcation of episodic selection. Mol. Biol. Evol. 32 , 1365–1371. https://doi.org/10.1093/molbev/msv035.

51. Kosakovsky Pond, S.L., and Frost, S.D.W. (2005). Not So Dierent After All: A Comparison

of Methods for Detecting Amino Acid Sites Under Selection. Mol. Biol. Evol. 22 , 1208–1222. https://doi.org/10.1093/molbev/msi105.

52. Murrell, B., Wertheim, J.O., Moola, S., Weighill, T., Scheer, K., and Kosakovsky Pond, S.L.

(2012). Detecting individual sites subject to episodic diversifying selection. PLoS Genet. 8 , e1002764. https://doi.org/10.1371/journal.pgen.1002764.

53. Blum, M., Chang, H.-Y., Chuguransky, S., Grego, T., Kandasaamy, S., Mitchell, A., Nuka, G.,

Paysan-Lafosse, T., Qureshi, M., Raj, S., et al. (2021). The InterPro protein families and domains database: 20 years on. Nucleic Acids Res. 49 , D344–D354. https://doi.org/10.1093/nar/gkaa977.

54. Brandes, N., Goldman, G., Wang, C.H., Ye, C.J., and Ntranos, V. (2023). Genome-wide

32

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 33

prediction of disease variant eects with a deep protein language model. Nat. Genet. 55 , 1512–1522. https://doi.org/10.1038/s41588-023-01465-0.

55. Perini, F.A., Russo, C. a. M., and Schrago, C.G. (2010). The evolution of South American

endemic canids: a history of rapid diversiﬁcation and morphological parallelism. J. Evol. Biol. 23 , 311–322. https://doi.org/10.1111/j.1420-9101.2009.01901.x.

56. Porto, L.M.V., Etienne, R.S., and Maestri, R. (2023). Evolutionary radiation in canids

following continental colonizations. Evolution 77 , 971–979. https://doi.org/10.1093/evolut/qpad015.

57. Frosali, S., Bartolini-Lucenti, S., Madurell-Malapeira, J., Urciuoli, A., Costeur, L., and Rook,

L. (2023). First digital study of the frontal sinus of stem-Canini (Canidae, Carnivora): evolutionary and ecological insights throughout advanced diagnostic in paleobiology. Front. Ecol. Evol. 11 , 1173341. https://doi.org/10.3389/fevo.2023.1173341.

58. Rook, L. (2009). The wide ranging genus Eucyon Tedford & Qiu, 1996 (Mammalia, Carnivora, Canidae, Canini) in the Mio-Pliocene of the Old World. Geodiversitas 31 , 723–741. https://doi.org/10.5252/g2009n4a723.

59. Werdelin, L., Lewis, M.E., and Haile‐Selassie, Y. (2015). A critical review of A frican species

of E ucyon ( M ammalia; C arnivora; C anidae), with a new species from the P liocene of the W oranso‐ M ille A rea, A far R egion, E thiopia. Pap. Palaeontol. 1 , 33–40. https://doi.org/10.1002/spp2.1001.

60. Bartolini Lucenti, S., and Rook, L. (2021). “Canis” ferox Revisited: Diet Ecomorphology of

Some Long Gone (Late Miocene and Pliocene) Fossil Dogs. J. Mamm. Evol. 28 , 285–306. https://doi.org/10.1007/s10914-020-09500-1.

61. Valenciano, A., Morales, J., and Govender, R. (2022). Eucyon khoikhoi sp. nov. (Carnivora: Canidae) from Langebaanweg ‘E’ Quarry (early Pliocene, South Africa): the most complete African canini from the Mio-Pliocene. Zool. J. Linn. Soc. 194 , 366–394. https://doi.org/10.1093/zoolinnean/zlab022.

62. Prassack, K.A., and Walkup, L.C. (2022). Maybe So, Maybe Not: Canis lepophagus at

Hagerman Fossil Beds National Monument, Idaho, USA. J. Mamm. Evol. 29 , 313–333. https://doi.org/10.1007/s10914-021-09591-4.

63. van der Valk, T., Pečnerová, P., Díez-Del-Molino, D., Bergström, A., Oppenheimer, J.,

Hartmann, S., Xenikoudakis, G., Thomas, J.A., Dehasque, M., Sağlıcan, E., et al. (2021). Million-year-old DNA sheds light on the genomic history of mammoths. Nature 591 , 265–269. https://doi.org/10.1038/s41586-021-03224-9.

64. Mallet, J. (2007). Hybrid speciation. Nature 446 , 279–283. https://doi.org/10.1038/nature05706.

65. Gopalakrishnan, S., Sinding, M.-H.S., Ramos-Madrigal, J., Niemann, J., Samaniego

33

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 34

Castruita, J.A., Vieira, F.G., Carøe, C., Montero, M.D.M., Kuderna, L., Serres, A., et al. (2018). Interspeciﬁc gene ﬂow shaped the evolution of the genus Canis. Curr. Biol. 28 , 3441-3449.e5. https://doi.org/10.1016/j.cub.2018.08.041.

66. Parmley, J.L., Chamary, J.V., and Hurst, L.D. (2006). Evidence for purifying selection

against synonymous mutations in mammalian exonic splicing enhancers. Mol. Biol. Evol. 23 , 301–309. https://doi.org/10.1093/molbev/msj035.

67. Lawrie, D.S., Messer, P.W., Hershberg, R., and Petrov, D.A. (2013). Strong purifying

selection at synonymous sites in D. melanogaster. PLoS Genet. 9 , e1003527. https://doi.org/10.1371/journal.pgen.1003527.

68. Gudkov, M., Thibaut, L., and Giannoulatou, E. (2024). Quantifying negative selection on

synonymous variants. Hum. Genet. Genomics Adv. 5 , 100262. https://doi.org/10.1016/j.xhgg.2024.100262.

69. Christmas, M.J., Dong, M., Meadows, J.R.S., Kozyrev, S.V., and Lindblad-Toh, K. (2024).

Interpreting mammalian evolutionary constraint at synonymous sites in light of the unwanted transcript hypothesis. Preprint, https://doi.org/10.1101/2024.04.23.590689 https://doi.org/10.1101/2024.04.23.590689.

70. Conway, E., Jerman, E., Healy, E., Ito, S., Holoch, D., Oliviero, G., Deevy, O., Glancy, E.,

Fitzpatrick, D.J., Mucha, M., et al. (2018). A Family of Vertebrate-Speciﬁc Polycombs Encoded by the LCOR/LCORL Genes Balance PRC2 Subtype Activities. Mol. Cell 70 , 408-421.e8. https://doi.org/10.1016/j.molcel.2018.03.005.

71. Plassais, J., Kim, J., Davis, B.W., Karyadi, D.M., Hogan, A.N., Harris, A.C., Decker, B.,

Parker, H.G., and Ostrander, E.A. (2019). Whole genome sequencing of canids reveals genomic regions under selection and variants inﬂuencing morphology. Nat. Commun. 10 , 1489. https://doi.org/10.1038/s41467-019-09373-w.

72. Graber, J.K., Signer‐Hasler, H., Burren, A., and Drögemüller, C. (2022). Evaluation of

truncating variants in the LCORL gene in relation to body size of goats from Switzerland. Anim. Genet. 53 , 237–239. https://doi.org/10.1111/age.13177.

73. Yuan, Z., Ge, L., Su, P., Gu, Y., Chen, W., Cao, X., Wang, S., Lv, X., Getachew, T., Mwacharo,

J.M., et al. (2023). NCAPG Regulates Myogenesis in Sheep, and SNPs Located in Its Putative Promoter Region Are Associated with Growth and Development Traits. Anim. Open Access J. MDPI 13 , 3173. https://doi.org/10.3390/ani13203173.

74. Metzger, J., Schrimpf, R., Philipp, U., and Distl, O. (2013). Expression levels of LCORL are

associated with body size in horses. PloS One 8 , e56497. https://doi.org/10.1371/journal.pone.0056497.

75. Majeres, L.E., Dilger, A.C., Shike, D.W., McCann, J.C., and Beever, J.E. (2024). Deﬁning a

Haplotype Encompassing the LCORL-NCAPG Locus Associated with Increased Lean Growth in Beef Cattle. Genes 15 , 576. https://doi.org/10.3390/genes15050576.

34

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 35

76. Bai, F., Cai, Y., Qiu, M., Liang, C., Pan, L., Liu, Y., Feng, Y., Cao, X., Yang, Q., Ren, G., et al.

(2025). LCORL and STC2 variants increase body size and growth rate in cattle and other animals. Preprint, https://doi.org/10.1101/2025.01.01.630985 https://doi.org/10.1101/2025.01.01.630985.

77. Zhang, W., Li, J., Guo, Y., Zhang, L., Xu, L., Gao, X., Zhu, B., Gao, H., Ni, H., and Chen, Y.

(2016). Multi-strategy genome-wide association studies identify the DCAF16-NCAPG region as a susceptibility locus for average daily gain in cattle. Sci. Rep. 6 , 38073. https://doi.org/10.1038/srep38073.

78. Jun, H.J., Roy, J., Smith, T.B., Wood, L.B., Lane, K., Woolfenden, S., Punko, D., Bronson, R.T.,

Haigis, K.M., Breton, S., et al. (2014). ROS1 signaling regulates epithelial dierentiation in the epididymis. Endocrinology 155 , 3661–3673. https://doi.org/2020071612140959800.

79. Basciani, S., Mariani, S., Arizzi, M., Brama, M., Ricci, A., Betsholtz, C., Bondjers, C., Ricci, G.,

Catizone, A., Galdieri, M., et al. (2004). Expression of platelet-derived growth factor (PDGF) in the epididymis and analysis of the epididymal development in PDGF-A, PDGF-B, and PDGF receptor beta deﬁcient mice. Biol. Reprod. 70 , 168–177. https://doi.org/10.1095/biolreprod.103.019232.

80. Schultz, N.G., Ingels, J., Hillhouse, A., Wardwell, K., Chang, P.L., Cheverud, J.M., Lutz, C.,

Lu, L., Williams, R.W., and Dean, M.D. (2016). The Genetic Basis of Baculum Size and Shape Variation in Mice. G3 GenesGenomesGenetics 6 , 1141–1151. https://doi.org/10.1534/g3.116.027888.

81. Peltzer, A., Jäger, G., Herbig, A., Seitz, A., Kniep, C., Krause, J., and Nieselt, K. (2016).

EAGER: ecient ancient genome reconstruction. Genome Biol. 17 , 60. https://doi.org/10.1186/s13059-016-0918-z.

82. Green, R.E., Malaspinas, A.-S., Krause, J., Briggs, A.W., Johnson, P.L.F., Uhler, C., Meyer, M.,

Good, J.M., Maricic, T., Stenzel, U., et al. (2008). A Complete Neandertal Mitochondrial Genome Sequence Determined by High-Throughput Sequencing. Cell 134 , 416–426. https://doi.org/10.1016/j.cell.2008.06.021.

83. Westbury, M.V., and Lorenzen, E.D. (2022). Iteratively mapping ancient DNA to reconstruct

highly divergent mitochondrial genomes: An evaluation of software, parameters and bait reference. Methods Ecol. Evol. 13 , 2419–2428. https://doi.org/10.1111/2041-210X.13990.

84. Durand, E.Y., Patterson, N., Reich, D., and Slatkin, M. (2011). Testing for Ancient Admixture

between Closely Related Populations. Mol. Biol. Evol. 28 , 2239–2252. https://doi.org/10.1093/molbev/msr048.

85. Gopalakrishnan, S., Ebenesersdóttir, S.S., Lundstrøm, I.K.C., Turner-Walker, G., Moore,

K.H.S., Luisi, P., Margaryan, A., Martin, M.D., Ellegaard, M.R., Magnússon, Ó.Þ., et al. (2022). The population genomic legacy of the second plague pandemic. Curr. Biol. CB 32 , 4743-4751.e6. https://doi.org/10.1016/j.cub.2022.09.023.

35

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 36

86. Sawyer, S., Krause, J., Guschanski, K., Savolainen, V., and Pääbo, S. (2012). Temporal

patterns of nucleotide misincorporations and DNA fragmentation in ancient DNA. PloS One 7 , e34131. https://doi.org/10.1371/journal.pone.0034131.

87. Zhang, X., Liu, T., Wang, J., Wang, P., Qiu, Y., Zhao, W., Pang, S., Li, X., Wang, H., Song, J.,

et al. (2021). Pan-genome of Raphanus highlights genetic variation and introgression among domesticated, wild, and weedy radishes. Mol. Plant 14 , 2032–2055. https://doi.org/10.1016/j.molp.2021.08.005.

88. Zhou, Y., Yang, L., Han, X., Han, J., Hu, Y., Li, F., Xia, H., Peng, L., Boschiero, C., Rosen, B.D.,

et al. (2022). Assembly of a pangenome for global cattle reveals missing sequences and novel structural variations, providing new insights into their diversity and evolutionary history. Genome Res. 32 , 1585–1601. https://doi.org/10.1101/gr.276550.122.

89. Martiniano, R., Garrison, E., Jones, E.R., Manica, A., and Durbin, R. (2020). Removing

reference bias and improving indel calling in ancient DNA data analysis by mapping to a sequence variation graph. Genome Biol. 21 , 250. https://doi.org/10.1186/s13059-020-02160-7.

90. Sirak, K., Fernandes, D., Cheronet, O., Harney, E., Mah, M., Mallick, S., Rohland, N.,

Adamski, N., Broomandkhoshbacht, N., Callan, K., et al. (2020). Human auditory ossicles as an alternative optimal source of ancient DNA. Genome Res. 30 , 427–436. https://doi.org/10.1101/gr.260141.119.

91. Ibrahim, J., Brumfeld, V., Addadi, Y., Rubin, S., Weiner, S., and Boaretto, E. (2022). The

petrous bone contains high concentrations of osteocytes: One possible reason why ancient DNA is better preserved in this bone. PLOS ONE 17 , e0269348. https://doi.org/10.1371/journal.pone.0269348.

92. Rohland, N., Glocke, I., Aximu-Petri, A., and Meyer, M. (2018). Extraction of highly

degraded DNA from ancient bones, teeth and sediments for high-throughput sequencing. Nat. Protoc. 13 , 2447–2461. https://doi.org/10.1038/s41596-018-0050-5.

93. Kapp, J.D., Green, R.E., and Shapiro, B. (2021). A Fast and Ecient Single-stranded

Genomic Library Preparation Method Optimized for Ancient DNA. J. Hered. 112 , 241–249. https://doi.org/10.1093/jhered/esab012.

94. Chen, S., Zhou, Y., Chen, Y., and Gu, J. (2018). fastp: an ultra-fast all-in-one FASTQ

preprocessor. Bioinformatics 34 , i884–i890. https://doi.org/10.1093/bioinformatics/bty560.

95. Wood, D.E., Lu, J., and Langmead, B. (2019). Improved metagenomic analysis with

Kraken 2. Genome Biol. 20 , 257. https://doi.org/10.1186/s13059-019-1891-0.

96. Lu, J., Rincon, N., Wood, D.E., Breitwieser, F.P., Pockrandt, C., Langmead, B., Salzberg, S.L.,

and Steinegger, M. (2022). Metagenome analysis using the Kraken software suite. Nat. Protoc. 17 , 2815–2839. https://doi.org/10.1038/s41596-022-00738-y.

36

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 37

97. Campana, M.G., Parker, L.D., Hawkins, M.T.R., Young, H.S., Helgen, K.M., Szykman Gunther,

M., Woodroe, R., Maldonado, J.E., and Fleischer, R.C. (2016). Genome sequence, population history, and pelage genetics of the endangered African wild dog ( Lycaon pictus ). BMC Genomics 17 , 1013. https://doi.org/10.1186/s12864-016-3368-9.

98. Kukekova, A.V., Johnson, J.L., Xiang, X., Feng, S., Liu, S., Rando, H.M., Kharlamova, A.V.,

Herbeck, Y., Serdyukova, N.A., Xiong, Z., et al. (2018). Red fox genome assembly identiﬁes genomic regions associated with tame and aggressive behaviours. Nat. Ecol. Evol. 2 , 1479–1491. https://doi.org/10.1038/s41559-018-0611-6.

99. Armstrong, E.E., Taylor, R.W., Prost, S., Blinston, P., vanderMeer, E., Madzikanda, H.,

Mufute, O., Mandisodza-Chikerema, R., Stuelpnagel, J., Sillero-Zubiri, C., et al. (2019). Cost-eective assembly of the African wild dog ( Lycaon pictus ) genome using linked reads. GigaScience 8 . https://doi.org/10.1093/gigascience/giy124.

100. Wang, G.-D., Shao, X.-J., Bai, B., Wang, J., Wang, X., Cao, X., Liu, Y.-H., Wang, X., Yin, T.-T.,

Zhang, S.-J., et al. (2019). Structural variation during dog domestication: insights from gray wolf and dhole genomes. Natl. Sci. Rev. 6 , 110–122. https://doi.org/10.1093/nsr/nwy076.

101. Allio, R., Tilak, M.-K., Scornavacca, C., Avenant, N.L., Kitchener, A.C., Corre, E., Nabholz, B.,

and Delsuc, F. (2021). High-quality carnivoran genomes from roadkill samples enable comparative species delineation in aardwolf and bat-eared fox. eLife 10 , e63167. https://doi.org/10.7554/eLife.63167.

102. Edwards, R.J., Field, M.A., Ferguson, J.M., Dudchenko, O., Keilwagen, J., Rosen, B.D.,

Johnson, G.S., Rice, E.S., Hillier, L.D., Hammond, J.M., et al. (2021). Chromosome-length genome assembly and structural variations of the primal Basenji dog ( Canis lupus familiaris ) genome. BMC Genomics 22 , 188. https://doi.org/10.1186/s12864-021-07493-6.

103. Halo, J.V., Pendleton, A.L., Shen, F., Doucet, A.J., Derrien, T., Hitte, C., Kirby, L.E., Myers, B.,

Sliwerska, E., Emery, S., et al. (2021). Long-read assembly of a Great Dane genome highlights the contribution of GC-rich sequence and mobile elements to canine genomes. Proc. Natl. Acad. Sci. 118 , e2016274118. https://doi.org/10.1073/pnas.2016274118.

104. Jagannathan, V., Hitte, C., Kidd, J.M., Masterson, P., Murphy, T.D., Emery, S., Davis, B.,

Buckley, R.M., Liu, Y.-H., Zhang, X.-Q., et al. (2021). Dog10K_Boxer_Tasha_1.0: a long-read assembly of the Dog reference genome. Genes 12 , 847. https://doi.org/10.3390/genes12060847.

105. Peng, Y., Li, H., Liu, Z., Zhang, C., Li, K., Gong, Y., Geng, L., Su, J., Guan, X., Liu, L., et al.

(2021). Chromosome‐level genome assembly of the Arctic fox ( Vulpes lagopus ) using PacBio sequencing and Hi‐C technology. Mol. Ecol. Resour. 21 , 2093–2108. https://doi.org/10.1111/1755-0998.13397.

106. Wang, C., Wallerman, O., Arendt, M.-L., Sundström, E., Karlsson, Å., Nordin, J., Mäkeläinen,

S., Pielberg, G.R., Hanson, J., Ohlsson, Å., et al. (2021). A novel canine reference genome

37

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 38

resolves genomic architecture and uncovers transcript complexity. Commun. Biol. 4 , 185. https://doi.org/10.1038/s42003-021-01698-x.

107. Chavez, D.E., Gronau, I., Hains, T., Dikow, R.B., Frandsen, P.B., Figueiró, H.V., Garcez, F.S.,

Tchaicka, L., De Paula, R.C., Rodrigues, F.H.G., et al. (2022). Comparative genomics uncovers the evolutionary history, demography, and molecular adaptations of South American canids. Proc. Natl. Acad. Sci. 119 , e2205986119. https://doi.org/10.1073/pnas.2205986119.

108. Field, M.A., Yadav, S., Dudchenko, O., Esvaran, M., Rosen, B.D., Skvortsova, K., Edwards,

R.J., Keilwagen, J., Cochran, B.J., Manandhar, B., et al. (2022). The Australian dingo is an early oshoot of modern breed dogs. Sci. Adv. 8 , eabm5944. https://doi.org/10.1126/sciadv.abm5944.

109. Lan, T., Li, H., Yang, S., Shi, M., Han, L., Sahu, S.K., Lu, Y., Wang, J., Zhou, M., Liu, H., et al.

(2022). The chromosome-scale genome of the raccoon dog: Insights into its evolutionary characteristics. iScience 25 , 105117. https://doi.org/10.1016/j.isci.2022.105117.

110. Lyu, T.-S., Wei, Q.-G., Wang, L.-D., Zhou, S.-Y., Shi, L.-P., Dong, Y.-H., Dou, H.-S., Sha, W.-L.,

Ga, T., Zhang, H.-H., et al. (2022). High-quality chromosome-level genome assembly of Tibetan fox ( Vulpes ferrilata ). Zool. Res. 43 , 362–366. https://doi.org/10.24272/j.issn.2095-8137.2021.399.

111. The Darwin Tree of Life Project Consortium, Blaxter, M., Mieszkowska, N., Di Palma, F.,

Holland, P., Durbin, R., Richards, T., Berriman, M., Kersey, P., Hollingsworth, P., et al. (2022). Sequence locally, think globally: The Darwin Tree of Life Project. Proc. Natl. Acad. Sci. 119 , e2115642118. https://doi.org/10.1073/pnas.2115642118.

112. Armstrong, E.E., Bissell, K.L., Heikkinen, M.A., Jessup, A., Junaid, M.O., Lee, D.H., Lieb, E.C.,

Liem, J.T., Martin, E.M., Moreno, M., et al. (2023). Chromosome-level assembly of the gray fox ( Urocyon cinereoargenteus ) conﬁrms the basal loss of PRDM9 in Canidae. Preprint at Genomics, https://doi.org/10.1101/2023.11.08.566296 https://doi.org/10.1101/2023.11.08.566296.

113. Bredemeyer, K.R., vonHoldt, B.M., Foley, N.M., Childers, I.R., Brzeski, K.E., and Murphy,

W.J. (2024). The value of hybrid genomes: Building two highly contiguous reference genome assemblies to advance Canis genomic studies. J. Hered. 115 , 480–486. https://doi.org/10.1093/jhered/esae013.

114. Kliver, S., Kovacic, I., Mak, S., Sinding, M.-H.S., Stagegaard, J., Petersen, B., Nesme, J.,

and Gilbert, M.T.P. (2024). A chromosome phased diploid genome assembly of African hunting dog ( Lycaon pictus ). J. Hered., esae052. https://doi.org/10.1093/jhered/esae052.

115. Sinding, M.-H.S., Gopalakrishan, S., Vieira, F.G., Samaniego Castruita, J.A., Raundrup, K.,

Heide Jørgensen, M.P., Meldgaard, M., Petersen, B., Sicheritz-Ponten, T., Mikkelsen, J.B., et al. (2018). Population genomics of grey wolves and wolf-like canids in North America.

38

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 39

PLOS Genet. 14 , e1007745. https://doi.org/10.1371/journal.pgen.1007745.

116. Li, H., and Durbin, R. (2009). Fast and accurate short read alignment with

Burrows–Wheeler transform. Bioinformatics 25 , 1754–1760. https://doi.org/10.1093/bioinformatics/btp324.

117. Oliva, A., Tobler, R., Llamas, B., and Souilmi, Y. (2021). Additional evaluations show that

speciﬁc BWA‐aln settings still outperform BWA‐mem for ancient DNA data alignment. Ecol. Evol. 11 , 18743–18748. https://doi.org/10.1002/ece3.8297.

118. Schubert, M., Ermini, L., Sarkissian, C.D., Jónsson, H., Ginolhac, A., Schaefer, R., Martin,

M.D., Fernández, R., Kircher, M., McCue, M., et al. (2014). Characterization of ancient and modern genomes by SNP detection and phylogenomic and metagenomic analysis using PALEOMIX. Nat. Protoc. 9 , 1056–1082. https://doi.org/10.1038/nprot.2014.063.

119. Jónsson, H., Ginolhac, A., Schubert, M., Johnson, P.L.F., and Orlando, L. (2013).

mapDamage2.0: fast approximate Bayesian estimates of ancient DNA damage parameters. Bioinformatics 29 , 1682–1684. https://doi.org/10.1093/bioinformatics/btt193.

120. Neukamm, J., Peltzer, A., and Nieselt, K. (2021). DamageProﬁler: fast damage pattern

calculation for ancient DNA. Bioinformatics 37 , 3652–3653. https://doi.org/10.1093/bioinformatics/btab190.

121. Pedersen, B.S., and Quinlan, A.R. (2018). Mosdepth: quick coverage calculation for

genomes and exomes. Bioinformatics 34 , 867–868. https://doi.org/10.1093/bioinformatics/btx699.

122. Okonechnikov, K., Conesa, A., and García-Alcalde, F. (2016). Qualimap 2: advanced

multi-sample quality control for high-throughput sequencing data. Bioinformatics 32 , 292–294. https://doi.org/10.1093/bioinformatics/btv566.

123. Wurster-Hill, D.H., Ward, O.G., Davis, B.H., Park, J.P., Moyzis, R.K., and Meyne, J. (1988).

Fragile sites, telomeric DNA sequences, B chromosomes, and DNA content in raccoon dogs, Nyctereutes procyonoides, with comparative notes on foxes, coyote, wolf, and raccoon. Cytogenet. Genome Res. 49 , 278–281. https://doi.org/10.1159/000132677.

124. Tiersch, T.R., Chandler, R.W., Wachtel, S.S., and Elias, S. (1989). Reference standards for

ﬂow cytometry and application in comparative studies of nuclear DNA content. Cytometry 10 , 706–710. https://doi.org/10.1002/cyto.990100606.

125. Smeds, L., Kojola, I., and Ellegren, H. (2019). The evolutionary history of grey wolf Y

chromosomes. Mol. Ecol. 28 , 2173–2191. https://doi.org/10.1111/mec.15054.

126. Hamilton, C.K., Favetta, L.A., Di Meo, G.P., Floriot, S., Perucatti, A., Peippo, J., Kantanen,

J., Eggen, A., Iannuzzi, L., and King, W.A. (2009). Copy Number Variation of Testis-Speciﬁc Protein, Y-Encoded (TSPY) in 14 Dierent Breeds of Cattle (Bos taurus) .

39

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 40

Sex. Dev. 3 , 205–213. https://doi.org/10.1159/000228721.

127. Xue, Y., and Tyler-Smith, C. (2011). An exceptional gene: evolution of the TSPY gene family

in humans and other great apes. Genes 2 , 36–47. https://doi.org/10.3390/genes2010036.

128. Quinlan, A.R., and Hall, I.M. (2010). BEDTools: a ﬂexible suite of utilities for comparing

genomic features. Bioinformatics 26 , 841–842. https://doi.org/10.1093/bioinformatics/btq033.

129. Hickey, G., Monlong, J., Ebler, J., Novak, A.M., Eizenga, J.M., Gao, Y., Human Pangenome

Reference Consortium, Abel, H.J., Antonacci-Fulton, L.L., Asri, M., et al. (2023). Pangenome graph construction from genome alignments with Minigraph-Cactus. Nat. Biotechnol. https://doi.org/10.1038/s41587-023-01793-w.

130. Liao, W.-W., Asri, M., Ebler, J., Doerr, D., Haukness, M., Hickey, G., Lu, S., Lucas, J.K.,

Monlong, J., Abel, H.J., et al. (2023). A draft human pangenome reference. Nature 617 , 312–324. https://doi.org/10.1038/s41586-023-05896-x.

131. Sirén, J., Monlong, J., Chang, X., Novak, A.M., Eizenga, J.M., Markello, C., Sibbesen, J.A.,

Hickey, G., Chang, P.-C., Carroll, A., et al. (2021). Pangenomics enables genotyping of known structural variants in 5202 diverse genomes. Science 374 , abg8871. https://doi.org/10.1126/science.abg8871.

132. Hickey, G., Heller, D., Monlong, J., Sibbesen, J.A., Sirén, J., Eizenga, J., Dawson, E.T.,

Garrison, E., Novak, A.M., and Paten, B. (2020). Genotyping structural variants in pangenome graphs using the vg toolkit. Genome Biol. 21 , 35. https://doi.org/10.1186/s13059-020-1941-7.

133. Walker, B.J., Abeel, T., Shea, T., Priest, M., Abouelliel, A., Sakthikumar, S., Cuomo, C.A.,

Zeng, Q., Wortman, J., Young, S.K., et al. (2014). Pilon: An Integrated Tool for Comprehensive Microbial Variant Detection and Genome Assembly Improvement. PLoS ONE 9 , e112963. https://doi.org/10.1371/journal.pone.0112963.

134. Korneliussen, T.S., Albrechtsen, A., and Nielsen, R. (2014). ANGSD: Analysis of Next

Generation Sequencing Data. BMC Bioinformatics 15 , 356. https://doi.org/10.1186/s12859-014-0356-4.

135. Barlow, A., Hartmann, S., Gonzalez, J., Hofreiter, M., and Paijmans, J.L.A. (2020).

Consensify: A Method for Generating Pseudohaploid Genome Sequences from Palaeogenomic Datasets with Reduced Error Rates. Genes 11 , 50. https://doi.org/10.3390/genes11010050.

136. Auwera, G.A.V. de, and O’Connor, B.D. (2020). Genomics in the cloud: using Docker, GATK,

and WDL in Terra First edition. (O’Reilly).

137. Troy, W., Damas, J., Titus, A.J., and Cantarel, B.L. (2023). FLAG: Find, Label, Annotate

Genomes, a fully automated tool for genome gene structural and functional annotation

40

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 41

of highly fragmented non-model species. Preprint, https://doi.org/10.1101/2023.07.14.548907 https://doi.org/10.1101/2023.07.14.548907.

138. Barlow, A., Hartmann, S., Gonzalez, J., Hofreiter, M., and Paijmans, J.L.A. (2020).

Consensify: A Method for Generating Pseudohaploid Genome Sequences from Palaeogenomic Datasets with Reduced Error Rates. Genes 11 , 50. https://doi.org/10.3390/genes11010050.

139. Li, H. (2018). Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics 34 , 3094–3100. https://doi.org/10.1093/bioinformatics/bty191.

140. Durbin, R., De Sanctis, B., and Blumer, M. (2023). Rotate: A command-line program to

rotate circular DNA sequences to start at a given position or string. Wellcome Open Res. 8 , 401. https://doi.org/10.12688/wellcomeopenres.19568.1.

141. Edgar, R.C. (2004). MUSCLE: multiple sequence alignment with high accuracy and high

throughput. Nucleic Acids Res. 32 , 1792–1797. https://doi.org/10.1093/nar/gkh340.

142. Stamatakis, A. (2014). RAxML version 8: a tool for phylogenetic analysis and

post-analysis of large phylogenies. Bioinformatics 30 , 1312–1313. https://doi.org/10.1093/bioinformatics/btu033.

143. Atickem, A., Stenseth, N.Chr., Drouilly, M., Bock, S., Roos, C., and Zinner, D. (2018). Deep

divergence among mitochondrial lineages in African jackals. Zool. Scr. 47 , 1–8. https://doi.org/10.1111/zsc.12257.

144. Hempel, E., Bibi, F., Faith, J.T., Brink, J.S., Kaltho, D.C., Kamminga, P., Paijmans, J.L.A.,

Westbury, M.V., Hofreiter, M., and Zachos, F.E. (2021). Identifying the true number of specimens of the extinct blue antelope (Hippotragus leucophaeus). Sci. Rep. 11 , 2100. https://doi.org/10.1038/s41598-020-80142-2.

145. Hempel, E., Faith, J.T., Preick, M., de Jager, D., Barish, S., Hartmann, S., Grau, J.H.,

Moodley, Y., Gedman, G., Pirovich, K.M., et al. (2024). Colonial-driven extinction of the blue antelope despite genomic adaptation to low population size. Curr. Biol. 34 , 2020-2029.e6. https://doi.org/10.1016/j.cub.2024.03.051.

146. Mirarab, S., Reaz, R., Bayzid, Md.S., Zimmermann, T., Swenson, M.S., and Warnow, T.

(2014). ASTRAL: genome-scale coalescent-based species tree estimation. Bioinformatics 30 , i541–i548. https://doi.org/10.1093/bioinformatics/btu462.

147. Zhang, C., Rabiee, M., Sayyari, E., and Mirarab, S. (2018). ASTRAL-III: polynomial time

species tree reconstruction from partially resolved gene trees. BMC Bioinformatics 19 , 153. https://doi.org/10.1186/s12859-018-2129-y.

148. Yang, Z. (2007). PAML 4: phylogenetic analysis by maximum likelihood. Mol. Biol. Evol. 24 , 1586–1591. https://doi.org/10.1093/molbev/msm088.

41

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

#### Page 42

149. Hasegawa, M., Kishino, H., and Yano, T. (1985). Dating of the human-ape splitting by a

molecular clock of mitochondrial DNA. J. Mol. Evol. 22 , 160–174. https://doi.org/10.1007/BF02101694.

150. Puttick, M.N. (2019). MCMCtreeR: functions to prepare MCMCtree analyses and visualize

posterior ages on trees. Bioinformatics 35 , 5321–5322. https://doi.org/10.1093/bioinformatics/btz554.

151. Yun, T., Li, H., Chang, P.-C., Lin, M.F., Carroll, A., and McLean, C.Y. (2021). Accurate,

scalable cohort variant calls using DeepVariant and GLnexus. Bioinformatics 36 , 5582–5589. https://doi.org/10.1093/bioinformatics/btaa1081.

152. Lin, M.F., Rodeh, O., Penn, J., Bai, X., Reid, J.G., Krasheninina, O., and Salerno, W.J. (2018).

GLnexus: joint variant calling for large cohort sequencing. Preprint, https://doi.org/10.1101/343970 https://doi.org/10.1101/343970.

153. Pertea, G., and Pertea, M. (2020). GFF Utilities: GRead and GCompare.

F1000Research 9 , 304. https://doi.org/10.12688/f1000research.23297.1.

154. Shen, W., Le, S., Li, Y., and Hu, F. (2016). SeqKit: A Cross-Platform and Ultrafast Toolkit for

FASTA/Q File Manipulation. PLOS ONE 11 , e0163962. https://doi.org/10.1371/journal.pone.0163962.

155. Muse, S., and Gaut, B. (1994). A likelihood approach for comparing synonymous and

nonsynonymous nucleotide substitution rates, with application to the chloroplast genome. Mol. Biol. Evol. https://doi.org/10.1093/oxfordjournals.molbev.a040152.

156. Villanueva-Cañas, J.L., Laurie, S., and Albà, M.M. (2013). Improving Genome-Wide Scans

of Positive Selection by Using Protein Isoforms of Similar Length. Genome Biol. Evol. 5 , 457–467. https://doi.org/10.1093/gbe/evt017.

157. Kolberg, L., Raudvere, U., Kuzmin, I., Adler, P., Vilo, J., and Peterson, H. (2023).

g:Proﬁler—interoperable web service for functional enrichment analysis and gene identiﬁer mapping (2023 update). Nucleic Acids Res. 51 , W207–W212. https://doi.org/10.1093/nar/gkad347.

42

.CC-BY-NC-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 11, 2025. ; https://doi.org/10.1101/2025.04.09.647074doi: bioRxiv preprint

---

Source: [Fermat's Library](https://fermatslibrary.com/) — https://fermatslibrary.com/s/on-the-ancestry-and-evolution-of-the-extinct-dire-wolf. More annotated papers: https://fermatslibrary.com/llms.txt
