CoChat
Sign Up
Cited by 10Genetic diversity and population structureGenetic and phenotypic traits in livestockGenomics and Phylogenetic Studies

Genomes of critically endangered saola are shaped by population structure and purging

Open access
Garcia-Erill, Genís, Liu, Shanlin, Le, Minh Duc, Hurley, Martha M., Nguyen, Hung Dinh, Nguyen, Dzung Quoc, Nguyen, Dzung Huy, Santander, Cindy G. +26 more·2025·Cell
DOI 10.1016/j.cell.2025.03.040PMC PMC12173715
Open this paper in CoChat — free accountOpen PDF

Abstract

The saola is one of the most elusive large mammals, standing at the brink of extinction. We constructed a reference genome and resequenced 26 saola individuals, confirming the saola as a basal member of the Bovini. Despite its small geographic range, we found that the saola is partitioned into two populations with high genetic differentiation ( F ST = 0.49). We estimate that these populations diverged and started declining 5,000–20,000 years ago, possibly due to climate changes and exacerbated by increasing human activities. The saola has long tracts without genomic diversity; however, most of these tracts are not shared by the two populations. Saolas carry a high genetic load, yet their gradual decline resulted in the purging of the most deleterious genetic variation. Finally, we find that combining the two populations, e.g., in an eventual captive breeding program, would mitigate the genetic load and increase the odds of species survival.

AI snapshot

Written from the paper’s full text.

TL;DR

The saola genome reveals two highly differentiated populations (FST = 0.49) that diverged and declined over the last 5,000–20,000 years, with long diversity-depleted tracts mostly not shared between them. Despite high genetic load, gradual decline allowed purging, and simulations indicate mixing northern and southern founders would improve captive-population viability.

Key findings

  • A de novo reference genome and 26 resequenced saola place the saola as a basal member of Bovini, diverging from the cattle/water buffalo lineage ca. 14 million years ago; the species is split into two populations with high genetic differentiation (FST = 0.47–0.49).
  • Demographic analyses infer divergence and decline beginning 5,000–20,000 years ago, a fastsimcoal divergence estimate of 7.5 kya, PSMC effective population size never above 5,000 in the last 10,000 years, and a recent stairway-plot decline to Ne < 10 diploid individuals.
  • Approximately 20%–40% of each saola genome is covered by long (>1 Mb) runs of homozygosity; many genomic windows have no detectable diversity, but most such depleted regions are not shared between the two populations, and combining populations increases diversity.
  • Saola carry a high relative amount of segregating deleterious variation, comparable to eastern gorillas and island foxes; simulations suggest purging removed almost two-thirds of the total genetic load present at the onset of decline, and mixed-founder captive scenarios had higher viability (e.g., 24 founders: 70–145 of 200 simulations survived).

Methods

  • The authors generated a de novo reference genome from a male saola and whole-genome resequencing data from 31 saola, retaining 26 individuals after removing samples with insufficient endogenous DNA and merging duplicate samples; conservative filtering yielded 628,905 common transversion SNPs (MAF > 5%) across the 26 samples.
  • Population genomic analyses included PCAngsd/NGSadmix, mtDNA phylogenetics, runs of homozygosity with PLINK, heterozygosity estimation, Hudson’s FST, PSMC, stairway plot, fastsimcoal2, and forward SLiM simulations; sensitive analyses used an 8-sample high-quality subset (>6× depth) and 3 samples >15×.
  • Variant calling and filtering excluded transition mutations to accommodate damaged DNA, and analyses used genotype likelihoods or genotype calls depending on sample quality.

Limitations

  • The paper states that sub-optimal DNA quality risks damage-driven miscoding lesions; the authors mitigated this by using higher-depth subsets and complementary methods, but the risk remains.
  • The paper states that forward-looking simulations omit biological complexity such as social structure or mating behavior, and that management recommendations depend on locating live animals or future de-extinction programs.
  • Evident from the text: sample size is small, sequencing depth, DNA damage, and error rates vary substantially across samples, metadata are incomplete for most samples, and cross-species heterozygosity comparisons are sensitive to filtering decisions.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1016/j.cell.2025.03.040

Cites

139 of 139 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

18 of 56 cited sources checked · 9 supported · 0 flagged · accuracy not yet measured

  • Findingown data

    The draft saola genome assembly had a scaffold N50 of 2.3 Mb.

    “produce a draft genome assembly with scaffold N50 of 2.3 Mb using Platanus.35”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [35] Kajitani R. Toshimoto K. Noguchi H. Toyoda A. Ogura Y. Okuno M. Yabana M. Harada M. Nagayasu E. Maruyama H. Efficient de novo assembly of highly heterozygous genomes from whole-genome shotgun short reads Genome Res. 24 2014 1384 1395 10.1101/gr.170720.113 24755901 PMC4120091 source
  • Findingown data

    The saola genome assembly contained 5,952 of 6,253 BUSCOs (95.2%) based on the Laurasiatheria database.

    “The assembly contained 5,952 out of 6,253 Benchmarking Universal Single-Copy Orthologs (BUSCO)35,36 based on the Laurasiatheria (orthoDB v9) database (95.2%)”
    Quote in the paperNot a substantial claim2 of 2 sources checked
    • cites [35] Kajitani R. Toshimoto K. Noguchi H. Toyoda A. Ogura Y. Okuno M. Yabana M. Harada M. Nagayasu E. Maruyama H. Efficient de novo assembly of highly heterozygous genomes from whole-genome shotgun short reads Genome Res. 24 2014 1384 1395 10.1101/gr.170720.113 24755901 PMC4120091 source

      The Results (Benchmarks overview) describe assembly evaluation using NG50 and fosmid or BAC validation, not BUSCO counts. Searches for the claimed counts, saola, BUSCO, and Laurasiatheria found no passage in this paper that addresses the reported 95.2% result. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
    • cites [36] Simão F.A. Waterhouse R.M. Ioannidis P. Kriventseva E.V. Zdobnov E.M. BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs Bioinformatics 31 2015 3210 3212 10.1093/bioinformatics/btv351 26059717 source
  • Findingown data

    Simulations showed an overall increase in realized genetic load during the saola population size decline.

    “Simulations showed an overall increase in realized genetic load (the genetic load that is expressed as negative fitness effect; Bertorelle et al.49) during the saola population size decline”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Bertorelle et al.49 Bertorelle G. Raffini F. Bosse M. Bortoluzzi C. Iannucci A. Trucchi E. Morales H.E. van Oosterhout C. Genetic load: genomic estimates and applications in non-model animals Nat. Rev. Genet. 23 2022 492 503 10.1038/s41576-022-00448-x 35136196 source
  • Interpretationcitation

    The number of common polymorphisms in saola is underestimated because all transition mutations, typically about two-thirds of all SNPs, were conservatively removed.

    “This underestimates the true number of common polymorphisms in the saola, as a large number of sites were conservatively removed to reduce errors, including all transition mutations, which typically account for approximately 2/3 of all SNPs.41”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [41] 1000 Genomes Project Consortium Abecasis G.R. Altshuler D. Auton A. Brooks L.D. Durbin R.M. Gibbs R.A. Hurles M.E. McVean G.A. A map of human genome variation from population-scale sequencing Nature 467 2010 1061 1073 10.1038/nature09534 20981092 PMC3042601 source

      The Results sections on variant discovery and mutation discuss human SNPs, but the passages found do not give a transition-to-transversion ratio or say that transitions account for approximately two-thirds of SNPs. Searches for “transition,” “transversion,” and the two-thirds figure also found no discussion of saola or the removal of its mutations to reduce errors. (LLM, conf=0.75)

      Experimental — accuracy not yet measured
  • Findingown data

    In blocks with discordant topologies, trees joining saola with water buffalo were overrepresented (27.9%) compared with those joining it with cattle (4.5%).

    “In the blocks with discordant topologies, there was an overrepresentation of trees joining the saola with the water buffalo (27.9%), while a much smaller minority placed it with the cattle (4.5%)”
    Quote in the paperNot a substantial claim
  • Classified as probably not a substantial claim. 35
  • May combine two claims. 52
  • The quote only partly states this claim. 1

  • Classified as probably not a substantial claim.

    Premisecitation

    The effects of gene flow depend, among other factors, on whether the demographic history of the source population has allowed for purging of most strongly deleterious recessive mutations.

    “The effects of gene flow depend, among other factors, on whether the demographic history of the source population has allowed for purging of most of the strongly deleterious recessive mutations.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [24] Kyriazis C.C. Wayne R.K. Lohmueller K.E. Strongly deleterious mutations are a primary determinant of extinction risk due to inbreeding depression Evol. Lett. 5 2021 33 47 10.1002/evl3.209 33552534 PMC7857301 source

  • May combine two claims.

    Premisecitation

    Long-term small population size tends to lead to accumulation of weakly and moderately deleterious variation due to reduced efficacy of selection, but also removes strongly deleterious variation through genetic purging.

    “long-term small population size tends to lead to the accumulation of weakly and moderately deleterious variation due to a reduced efficacy of selection,16,17 but it also removes strongly deleterious variations through genetic purging.18,19”
    Quote in the paperSubstantial claim1 of 4 sources checked
    • cites [16] Ohta T. The Nearly Neutral Theory of Molecular Evolution Annu. Rev. Ecol. Syst. 23 1992 263 286 10.1146/annurev.es.23.110192.001403 source
    • cites [17] Leroy T. Rousselle M. Tilak M.-K. Caizergues A.E. Scornavacca C. Recuerda M. Fuchs J. Illera J.C. De Swardt D.H. Blanco G. Island songbirds as windows into evolution in small populations Curr. Biol. 31 2021 1303 1310.e4 10.1016/j.cub.2020.12.040 33476557 source
    • cites [18] Kirkpatrick M. Jarne P. The Effects of a Bottleneck on Inbreeding Depression and the Genetic Load Am. Nat. 155 2000 154 167 10.1086/303312 10686158 source
    • cites [19] Xue Y. Prado-Martinez J. Sudmant P.H. Narasimhan V. Ayub Q. Szpak M. Frandsen P. Chen Y. Yngvadottir B. Cooper D.N. Mountain gorilla genomes reveal the impact of long-term population decline and inbreeding Science 348 2015 242 245 10.1126/science.aaa3952 25859046 PMC4668944 source

  • May combine two claims.

    Premisecitation

    Small populations are vulnerable to genomic erosion due to genetic drift and inbreeding, leading to loss of genetic diversity and an increase in harmful mutations, threatening their long-term survival.

    “Small populations are vulnerable to genomic erosion due to genetic drift and inbreeding, which lead to a loss of genetic diversity and an increase in harmful mutations, threatening their long-term survival.”
    Quote in the paperSubstantial claim1 of 4 sources checked
    • cites [10] Caughley G. Directions in Conservation Biology J. Animal Ecol. 63 1994 215 244 10.2307/5542 source
    • cites [11] Bijlsma R. Loeschcke V. Genetic erosion impedes adaptive responses to stressful environments Evol. Appl. 5 2012 117 129 10.1111/j.1752-4571.2011.00214.x 25568035 PMC3353342 source
    • cites [12] Díez-Del-Molino D. Sánchez-Barreiro F. Barnes I. Gilbert M.T.P. Dalén L. Quantifying Temporal Genomic Erosion in Endangered Species Trends Ecol. Evol. 33 2018 176 185 10.1016/j.tree.2017.12.002 29289355 source
    • cites [13] van Oosterhout C. Speak S.A. Birley T. Bortoluzzi C. Percival-Alwyn L. Urban L.H. Groombridge J.J. Segelbacher G. Morales H.E. Genomic erosion in the assessment of species extinction risk and recovery potential Preprint at bioRxiv 2022 10.1101/2022.09.13.507768 PMC13539449 41626739 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Gene flow can also increase extinction risk by introducing strongly deleterious mutations.

    “However, gene flow can also increase extinction risk by introducing strongly deleterious mutations.”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [22] Hedrick P.W. Robinson J.A. Peterson R.O. Vucetich J.A. Genetics and extinction and the example of Isle Royale wolves Animal Conservation 22 2019 302 309 10.1111/acv.12479 source
    • cites [23] Robinson J.A. Räikkönen J. Vucetich L.M. Vucetich J.A. Peterson R.O. Lohmueller K.E. Wayne R.K. Genomic signatures of extensive inbreeding in Isle Royale wolves, a population on the threshold of extinction Sci. Adv. 5 2019 eaau0757 10.1126/sciadv.aau0757 PMC6541468 31149628 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Nearly 49% of the saola genome is composed of repetitive elements, similar to cattle (48%), with LINEs making up 26.22%.

    “Nearly 49% of the genome is composed of repetitive elements, similar to that of cattle (48%),39 the majority of which are long interspersed nuclear elements (LINEs; 26.22%)”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [39] Elsik C.G. Tellam R.L. Worley K.C. Gibbs R.A. Muzny D.M. Weinstock G.M. Adelson D.L. Eichler E.E. Elnitski L. Bovine Genome Sequencing Analysis Consortium The genome sequence of taurine cattle: a window to ruminant biology and evolution Science 324 2009 522 528 10.1126/science.1169588 19390049 PMC2943200 source

      The citation is attached to the cattle comparison, specifically the claim that repetitive elements make up 48% of the cattle genome. The repeat discussion in our copy (section labelled Abstract; page unavailable) says ruminant-specific repeats comprise 27%, but gives no percentage for all cattle repetitive elements. Searches for 48% and total repeat content did not verify the cited figure; the 27% figure describes a subset, not a contrary total. (LLM, conf=0.74)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingcitation

    The GC content of the saola coding region is 55.8%, higher than the whole genome (41.9%), similar to other mammals.

    “At 55.8%, the GC content of the coding region is considerably higher than that of the whole genome (41.9%), similar to reports of the GC distribution in other mammals.40”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [40] Jiang Y. Xie M. Chen W. Talbot R. Maddox J.F. Faraut T. Wu C. Muzny D.M. Li Y. Zhang W. The sheep genome illuminates biology of the rumen and lipid metabolism Science 344 2014 1168 1173 10.1126/science.1252806 24904168 PMC4157056 source

      The cited paper’s genome-assembly and comparative-genomics passages (section labeled Abstract; page unavailable) do not report GC content for coding regions or whole genomes. Searches for the stated percentages and for a mammalian coding-region GC comparison found no passage supporting the comparison attributed to this source. (LLM, conf=0.76)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Study design and conductcitation

    All time intervals from 2 to 500 kya have more than 100 recombinations, which is above the minimum of 10 recommended for reliable PSMC inference.

    “All time intervals from 2 to 500 kya have more than 100 recombinations, which is above the minimum of 10 recommended for reliable PSMC inference.46”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [46] Li H. Durbin R. Inference of human population history from individual whole-genome sequences Nature 475 2011 493 496 10.1038/nature10231 21753753 PMC3154645 source

      The model-validation passage (page unavailable) supports the general point that recombination events affect PSMC reliability. It does not give a minimum of 10 recombinations or report more than 100 in every interval from 2 to 500 kya; instead, it says estimates more recent than 20 kya have reduced accuracy and large variance. (LLM, conf=0.85)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Interpretationcitation

    The relatively recent inferred onset of divergence and population decline at ≈20 kya coincides with likely changes in the distribution of dense forest habitat during and after the LGM, suggesting that these major events in saola population history could be driven by habitat fragmentation and isolation into two geographically distinct areas.

    “The relatively recent inferred onset of divergence and population decline at ≈20 kya coincides with likely changes in the distribution of dense forest habitat during and after the LGM,51,52,53,54 suggesting that these major events in saola population history could be driven by habitat fragmentation and isolation into two geographically distinct areas.”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites [51] Wurster C.M. Bird M.I. Bull I.D. Creed F. Bryant C. Dungait J.A.J. Paz V. Forest contraction in north equatorial Southeast Asia during the Last Glacial Period Proc. Natl. Acad. Sci. USA 107 2010 15508 15511 10.1073/pnas.1005507107 20660748 PMC2932586 source
    • cites [52] Suraprasit K. Shoocongdej R. Chintakanon K. Bocherens H. Late Pleistocene human paleoecology in the highland savanna ecosystem of mainland Southeast Asia Sci. Rep. 11 2021 16756 10.1038/s41598-021-96260-4 PMC8373907 34408215 source
    • cites [53] Morley R.J. Morley H.P. 1 - The prelude to the Holocene: tropical Asia during the Pleistocene Kumaran N. Damodara P. Holocene Climate Change and Environment 2022 Elsevier 1 32 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). The quote only partly states this claim.

    Premisecitation

    The finding that a gradual population decline allowed purging is in line with previous research on small and endangered populations such as mountain gorillas, island foxes, vaquita, kākāpō, and the Ethiopian wolf.

    “This is in line with previous research on small and endangered populations, e.g., mountain gorillas,19,58 island foxes,59 vaquita,60,61 kākāpō,62 and the Ethiopian wolf.63”
    Quote in the paperNot a substantial claim2 of 3 sources checked
    • cites [19] Xue Y. Prado-Martinez J. Sudmant P.H. Narasimhan V. Ayub Q. Szpak M. Frandsen P. Chen Y. Yngvadottir B. Cooper D.N. Mountain gorilla genomes reveal the impact of long-term population decline and inbreeding Science 348 2015 242 245 10.1126/science.aaa3952 25859046 PMC4668944 source

      The Abstract reports a prolonged decline in eastern gorillas and evidence of purging severely deleterious mutations in mountain gorillas. It attributes the purging more specifically to extensive inbreeding following a further recent decline, rather than establishing that gradual decline itself allowed purging. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
    • cites [58] van der Valk T. Díez-Del-Molino D. Marques-Bonet T. Guschanski K. Dalén L. Historical Genomes Reveal the Genomic Consequences of Recent Population Decline in Eastern Gorillas Curr. Biol. 29 2019 165 170.e6 10.1016/j.cub.2018.11.055 30595519 source
    • cites [59] Robinson J.A. Brown C. Kim B.Y. Lohmueller K.E. Wayne R.K. Purging of Strongly Deleterious Mutations Explains Long-Term Persistence and Absence of Inbreeding Depression in Island Foxes Curr. Biol. 28 2018 3487 3494.e4 10.1016/j.cub.2018.08.066 30415705 PMC6462144 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Similar fragmentation has been invoked to explain the phylogeography of co-distributed species such as the large-antlered muntjac.

    “Similar fragmentation has also been invoked to explain the phylogeography of co-distributed species such as the large-antlered muntjac.55”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [55] Stimpson C.M. Utting B. O’Donnell S. Huong N.T.M. Kahlert T. Manh B.V. Khanh P.S. Rabett R.J. An 11 000-year-old giant muntjac subfossil from Northern Vietnam: implications for past and present populations R. Soc. Open Sci. 6 2019 181461 10.1098/rsos.181461 PMC6458398 31032005 source

      The Conclusion (page unavailable) says giant muntjacs were formerly more widely distributed and that remaining Annamite populations are refugial and subject to anthropogenic pressures. That supports a history of range contraction, but the paper does not invoke fragmentation to explain the species’ phylogeography. (LLM, conf=0.79)

      Experimental — accuracy not yet measured

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning

    Conservation strategies depend on information about past demographic history, and simulations of load are a vital tool for assessing genetic rescue and other conservation regimes.

    “Hence, we highlight that conservation strategies depend on information about past demographic history, and we demonstrate how simulations of load are a vital tool for assessing genetic rescue and other conservation regimes.62,76”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [62] Dussex N. van der Valk T. Morales H.E. Wheat C.W. Díez-Del-Molino D. von Seth J. Foster Y. Kutschera V.E. Guschanski K. Rhie A. Population genomics of the critically endangered kākāpō Cell Genom. 1 2021 100002 10.1016/j.xgen.2021.100002 PMC9903828 36777713 source

      The Methods (no page provided) describe simulations of mutational load under contrasting demographic histories, and the Discussion connects mutational load to conservation decisions, including genetic rescue. The paper does not describe simulations that assess genetic rescue or compare other conservation regimes directly, so that part of the attributed claim goes beyond its findings. (LLM, conf=0.85)

      Experimental — accuracy not yet measured
    • cites [76] Femerling G. van Oosterhout C. Feng S. Bristol R.M. Zhang G. Groombridge J. P Gilbert M.T. Morales H.E. Genetic Load and Adaptive Potential of a Recovered Avian Species that Narrowly Avoided Extinction Mol. Biol. Evol. 40 2023 msad256 10.1093/molbev/msad256 PMC10701096 37995319 source

      The Discussion section, ‘The Role of Genomics in Species Conservation Assessments’ (page unavailable), says that reconstructing past demography and combining those insights with simulations can inform conservation assessments. The Results section, ‘Individual-Based Simulations’ (page unavailable), models genetic load across different ancestral population sizes. Our copy does not show simulations being used to assess genetic rescue or compare other conservation regimes, so that specific application is unsupported. (LLM, conf=0.84)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    The dominant phylogenetic tree topology placing saola as sister to cattle and water buffalo was supported by an average of 67.9% of subtrees, with 82.0% on the X chromosome.

    “The dominant tree topology was supported by an average of 67.9% of subtrees, with relatively low variability across the autosomes but with the X chromosome showing an elevated proportion (82.0%) of subtrees supporting the species tree.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The saola has long tracts without genomic diversity, and most of these tracts are not shared by the two populations.

    “The saola has long tracts without genomic diversity; however, most of these tracts are not shared by the two populations.”
    • Also stated in: § Saola populations have retained complementary genetic diversity along the genome during their decline · hedged
      “In most regions where both populations had no diversity, there was diversity in the combined population, indicating that each population was fixed for a different haplotype”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The two saola populations diverged and started declining 5,000-20,000 years ago.

    “We estimate that these populations diverged and started declining 5,000–20,000 years ago”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The saola is a sister taxon to the cattle and water buffalo lineages, a basal member of the Bovini.

    “confirming the saola as a basal member of the Bovini”
    • Also stated in: § Phylogenomic analysis confirms the saola as a sister taxon to cattle and buffalo · plain
      “placed the saola as a sister group to the cattle and water buffalo lineages”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The saola is partitioned into two genetically differentiated populations, northern and southern, with high genetic differentiation (FST = 0.47-0.49 depending on analysis).

    “we found that the saola is partitioned into two populations with high genetic differentiation (FST = 0.49)”
    • Also stated in: § Two genetically differentiated saola populations with low genetic diversity and elevated proportion of deleterious genetic variation · plain
      “which results in an FST estimate of 0.47 and 0.49 depending on sample inclusion, mutation filtering criteria, and allele frequency inference methodologies”
    • Also stated in: § Two genetically differentiated saola populations with low genetic diversity and elevated proportion of deleterious genetic variation · boosted
      “consistently split the samples into two clearly differentiated genetic groups concordant with their geographic origins and without evidence of further discrete substructure within them”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Saolas carry a high genetic load, with a high relative amount of potentially deleterious segregating variation.

    “Saolas carry a high genetic load”
    • Also stated in: § Two genetically differentiated saola populations with low genetic diversity and elevated proportion of deleterious genetic variation · plain
      “we observed a high relative amount of potentially deleterious variation segregating in saola, comparable only to estimates for eastern gorillas and island foxes”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The decline was an abrupt decrease followed by a period of slow, but continuous, decline in population size.

    “and furthermore revealed that it was an abrupt decrease followed by a period of slow, but continuous, decline in population size (Figure 5B).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The saola genome annotation identified 21,054 protein-coding genes and 7,512 non-coding functional RNAs.

    “we annotated a total of 21,054 protein-coding genes (Table S3) and 7,512 non-coding functional RNAs.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Divergence was accompanied by a 17-fold decline in the northern population size and a 24-fold decline in the southern population size.

    “the demographic model suggested that divergence was accompanied by a 17- and 24-fold decline in the northern and southern population sizes, respectively.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Allele frequency-based estimates of inbreeding and kinship revealed some related pairs of individuals and four samples (three northern, one southern) with elevated inbreeding, indicative of being offspring of close relatives.

    “Allele frequency-based estimates of inbreeding and kinship revealed some related pairs of individuals and four samples (three in the northern and one in the southern) with elevated inbreeding (Figure S4D), indicative of being the offspring of close relatives.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The saola showed remarkably low heterozygosity, at the lower end of previously published estimates from other species, comparable with some other bovids believed to have been long restricted to small geographical areas.

    “The saola showed remarkably low heterozygosity, at the lower end of previously published estimates from other species, and comparable with some other bovids believed to have been long restricted to small geographical areas”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The last decline roughly coincides with the last glacial maximum (LGM) and is accompanied by a divergence of the Ne curve between the two populations.

    “This last decline roughly coincides with the last glacial maximum (LGM) and is accompanied by a divergence of the Ne curve between the two populations (Figure 5A).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In both saola populations, many genomic windows showed no detectable genetic diversity, with the southern population exhibiting a higher number of depleted regions.

    “In both of the populations, many windows across the genome showed no detectable genetic diversity, with the southern population exhibiting a higher number of depleted regions”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    After reference-assisted chromosome assembly, the saola genome assembly comprised 117 predicted chromosome fragments, assembled 90% of the total genome (2.5 G), and had a scaffold N50 of ca. 79 MB.

    “we were able to generate a total of 117 predicted chromosome fragments (PCFs), assembling 90% of the total genome (2.5 G) and improving the scaffold N50 to ca. 79 MB”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [38] Kim J. Larkin D.M. Cai Q. Asan Z. Zhang Y. Ge R.-L. Auvil L. Capitanu B. Zhang G. Lewin H.A. Reference-assisted chromosome assembly Proc. Natl. Acad. Sci. USA 110 2013 1785 1790 10.1073/pnas.1220349110 23307812 PMC3562798 source

  • May combine two claims.

    Findingown data

    Despite the realized load increasing in the succeeding generations, on average it does not go above the realized genetic load in the single population source scenarios, and in the last generation tends to be equal to or lower than.

    “Despite the realized load increasing in the succeeding generations, on average it does not go above the realized genetic load in the single population source scenarios, and in the last generation tends to be equal to or lower than.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    There is a consistent but weak positive correlation between the fraction of missing sites and diversity for the analysis of 100 kb windows.

    “There is a consistent but weak positive correlation.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Southern saola samples tended to have lower heterozygosity levels than northern samples, although the range of heterozygosity values for the two populations overlaps when ROHs are included.

    “Samples from the southern population tended to have lower heterozygosity levels than those from the northern population, although the range of heterozygosity values for the two populations overlaps when ROHs are included”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    After strict filtering, the population genomic dataset retained approximately 1.2 Gb with relatively uniform distribution across the genome, within which 628,905 common transversion SNPs (MAF > 5%) were identified across 26 saola samples.

    “we retained a total of ≈1.2 Gb with a relatively uniform distribution across the genome (Figure S2E) for population genetic analyses, within which we identified 628,905 common (minor allele frequency [MAF] > 5%) transversion SNPs across the 26 saola samples.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    PCA, admixture analysis, and mtDNA tree consistently split the saola samples into two clearly differentiated genetic groups concordant with geographic origins, without evidence of further discrete substructure within them.

    “A principal-component analysis, admixture analysis, and an mtDNA tree consistently split the samples into two clearly differentiated genetic groups concordant with their geographic origins and without evidence of further discrete substructure within them”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Cross-coalescence between the two saola populations declined, coinciding with the visible separation of population size trajectories during the LGM.

    “which revealed a decline in cross-coalescence between the populations coinciding with the visible separation of the population size trajectories during the LGM”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Genetic diversity in the combined saola population was higher, and the proportion of windows without detectable diversity was lower, than in either population alone.

    “Genetic diversity in the combined population was higher, and the proportion of windows without detectable diversity was lower”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Windows with more missingness are more likely to be depleted, but the bias is stronger in the totally depleted category.

    “There is a bias where windows with more missingness are more likely to be depleted, but the bias is stronger in the totally depleted category.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    For the mitochondrial alignment of 17,109 sites, GTR+I+G4 was selected as the best-fitting substitution scheme based on the Bayesian Information Criterion.

    “The alignment of 17,109 sites was used in ModelTest-NG v0.1.7106 to determine the best-fitting substitution scheme and GTR+I+G4 was selected based on the Bayesian Information Criterion.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Sample NVq9 was inferred to be a first-degree relative of other included samples.

    “after excluding a sample inferred to be first degree relative of other included samples (NVq9, see Figure S4D)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    A total of 644,546 transversion SNPs were called for 26 saola.

    “This resulted in a total of 644,546 transversion SNPs called for 26 saola.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    PSMC suggests very similar population size trajectories for both populations, with the Southern population having approximately two-thirds lower population sizes.

    “This was based on the fact that PSMC suggests very similar trajectories for both populations, with the Southern having approximately ⅔ lower population sizes.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Combining the two saola populations, for example in an eventual captive breeding program, would mitigate the genetic load and increase the odds of species survival.

    “we find that combining the two populations, e.g., in an eventual captive breeding program, would mitigate the genetic load and increase the odds of species survival”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Gene flow from external populations can improve population health by increasing genetic diversity and reducing segregation of deleterious variation, known as genetic rescue.

    “gene flow from external populations can improve population health by increasing genetic diversity and reducing the segregation of deleterious variation in what is known as genetic rescue.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [20] Miller J.M. Poissant J. Hogg J.T. Coltman D.W. Genomic consequences of genetic rescue in an insular population of bighorn sheep (Ovis canadensis) Mol. Ecol. 21 2012 1583 1596 10.1111/j.1365-294X.2011.05427.x 22257293 source
    • cites [21] Whiteley A.R. Fitzpatrick S.W. Funk W.C. Tallmon D.A. Genetic rescue to the rescue Trends Ecol. Evol. 30 2015 42 49 10.1016/j.tree.2014.10.009 25435267 source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    The decline of saola populations may have been due to climate changes and exacerbated by increasing human activities.

    “possibly due to climate changes and exacerbated by increasing human activities”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    Mixed captive breeding programs would not only increase the odds of short-term species survival but also increase long-term evolutionary potential in the resulting population.

    “emphasizing that mixed captive breeding programs would not only increase the odds of short-term species survival but also increase long-term evolutionary potential in the resulting population.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Although realized load increased over the last 10,000 generations in the saola, purging removed much of the genetic load that the species would have otherwise accumulated.

    “Hence, although we show that realized load increased over the last 10,000 generations in the saola, purging removed much of the genetic load that the species would have otherwise accumulated.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Sample N??1 clustered with northern samples despite a recorded southern origin; it was likely moved from its origin before collection rather than being a migrant, and was grouped with the northern population.

    “we consider it much more likely that the sample material was moved from its origin before being collected rather than it being a migrant, and we therefore grouped it with the northern population in all subsequent analyses.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    The lower coding-sequence proportion in ROH-accumulating regions may result from higher frequency of recessive deleterious variants in coding regions; individuals with ROH in coding-rich areas likely experience reduced fitness, promoting stronger genetic purging, so inbreeding tracts preferentially accumulate in genomic regions with less coding sequence.

    “This pattern may result from the higher frequency of recessive deleterious variants in coding regions. Individuals with ROH in areas rich in coding regions are likely to experience reduced fitness, which promotes stronger genetic purging in these regions. Thereby, inbreeding tracts will preferentially accumulate in genomic regions with less coding sequence.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    The overall conclusion regarding the dynamics of genetic load and purging is robust to modeling assumptions, as the two different models of selection and dominance coefficient distributions yielded different absolute values of genetic load but highly similar relative dynamics.

    “The two different models of selection and dominance coefficient distributions yielded different absolute values of genetic load, but the relative dynamics are highly similar, and the overall conclusion regarding the dynamics of genetic load and purging is robust to modeling assumptions (Figures S8B and S8C).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The inferred divergence time of 7.5 kya is roughly consistent with the time when PSMC relative cross-coalescence rates reach their minimum but not the time when they start decreasing.

    “which is roughly consistent with the time when PSMC relative cross-coalescence rates reach their minimum but not the time when they start decreasing (Figure 5A).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    The saola's phylogenetic placement within the Bovini differs among analyses, having been placed as sister to Bovina, to Bubalina, or as a basal branch within Bovini.

    “While cytogenetic and phylogenetic studies based on mtDNA and 13 introns have determined it is part of the tribe Bovini, its placement within the tribe differs in various analyses that have placed it inconsistently as a sister taxon to the Bovina,28,29,30 the Bubalina,31 or indeed as a basal branch within the Bovini tribe.32”
    Quote in the paperSubstantial claim0 of 5 sources checked
    • cites [28] Gatesy J. Arctander P. Hidden morphological support for the phylogenetic placement of Pseudoryx nghetinhensis with bovine bovids: a combined analysis of gross anatomical evidence and DNA sequences from five genes Syst. Biol. 49 2000 515 538 10.1080/10635159950127376 12116425 source
    • cites [29] Hassanin A. Delsuc F. Ropiquet A. Hammer C. Jansen van Vuuren B. Matthee C. Ruiz-Garcia M. Catzeflis F. Areskoug V. Nguyen T.T. Pattern and timing of diversification of Cetartiodactyla (Mammalia, Laurasiatheria), as revealed by a comprehensive analysis of mitochondrial genomes C. R. Biol. 335 2012 32 50 10.1016/j.crvi.2011.11.002 22226162 source
    • cites [30] Hassanin A. Douzery E.J. Evolutionary affinities of the enigmatic saola (Pseudoryx nghetinhensis) in the context of the molecular phylogeny of Bovidae Proc. Biol. Sci. 266 1999 893 900 10.1098/rspb.1999.0720 10380679 PMC1689916 source
    • cites [31] Nguyen T.T. Aniskin V.M. Gerbault-Seureau M. Planton H. Renard J.P. Nguyen B.X. Hassanin A. Volobouev V.T. Phylogenetic position of the saola (Pseudoryx nghetinhensis) inferred from cytogenetic analysis of eleven species of Bovidae Cytogenet. Genome Res. 122 2008 41 54 10.1159/000151315 18931485 source
    • cites [32] Hassanin A. An J. Ropiquet A. Nguyen T.T. Couloux A. Combining multiple autosomal introns for studying shallow phylogeny and taxonomy of Laurasiatherian mammals: Application to the tribe Bovini (Cetartiodactyla, Bovidae) Mol. Phylogenet. Evol. 66 2013 766 775 10.1016/j.ympev.2012.11.003 23159894 source

  • May combine two claims.

    Study design and conductown data

    The pattern of complementary diversity between saola populations was consistent across window sizes from 50 to 500 kbp and was not driven by bias from missing sites in genetic diversity estimation.

    “The pattern was consistent across window sizes from 50 to 500 kbp (Figure S6A) and was not driven by the bias introduced by missing sites in the estimation of genetic diversity”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductnone given

    The available saola samples contained sub-optimal quality DNA, meaning there is a risk that damage-driven miscoding lesions may affect the results.

    “The fact that the available saola samples contained sub-optimal quality DNA means there is a risk that damage-driven miscoding lesions may affect our results.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning

    Managing the saola as a single population would increase the saola's long-term survival chances by both reducing genetic load and increasing the amount of standing genetic variation available to adapt to future environmental changes.

    “This would increase the saola’s long-term survival chances by both reducing genetic load and increasing the amount of standing genetic variation available to adapt to future environmental changes.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductcitation

    For the population-level analyses, a combination of different approaches was used to identify and exclude problematic regions in the reference genome that are more likely to contain mapping or genotyping errors.

    “For the population-level analyses we used a combination of different approaches to identify and exclude problematic regions in the reference genome that are more likely to contain mapping or genotyping errors.133”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Wherever possible, results were corroborated by using alternative and/or complementary methods, e.g., in the heterozygosity and FST estimation.

    “Wherever possible, we corroborated results by using alternative and/or complementary methods, e.g., in the heterozygosity and FST estimation.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    Whole-genome data reject the phylogenetic positions inferred through previous analyses based upon cytogenetics and single-marker mitochondrial genomes, and uphold the topology previously inferred from 13 intron regions.

    “Hence, using the increased power of whole-genome data, we reject the phylogenetic positions inferred through previous analyses based upon cytogenetics31 and single-marker mitochondrial genomes29 and uphold the topology previously inferred from 13 intron regions,32 and consistent with the assessment of some previous studies.50”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [31] Nguyen T.T. Aniskin V.M. Gerbault-Seureau M. Planton H. Renard J.P. Nguyen B.X. Hassanin A. Volobouev V.T. Phylogenetic position of the saola (Pseudoryx nghetinhensis) inferred from cytogenetic analysis of eleven species of Bovidae Cytogenet. Genome Res. 122 2008 41 54 10.1159/000151315 18931485 source
    • cites [29] Hassanin A. Delsuc F. Ropiquet A. Hammer C. Jansen van Vuuren B. Matthee C. Ruiz-Garcia M. Catzeflis F. Areskoug V. Nguyen T.T. Pattern and timing of diversification of Cetartiodactyla (Mammalia, Laurasiatheria), as revealed by a comprehensive analysis of mitochondrial genomes C. R. Biol. 335 2012 32 50 10.1016/j.crvi.2011.11.002 22226162 source
    • cites [32] Hassanin A. An J. Ropiquet A. Nguyen T.T. Couloux A. Combining multiple autosomal introns for studying shallow phylogeny and taxonomy of Laurasiatherian mammals: Application to the tribe Bovini (Cetartiodactyla, Bovidae) Mol. Phylogenet. Evol. 66 2013 766 775 10.1016/j.ympev.2012.11.003 23159894 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    The forward-looking simulations do not incorporate all relevant elements of biological complexity, such as social structure or mating behavior, because no relevant data are available to parameterize these.

    “We additionally acknowledge that our forward-looking simulations do not incorporate all relevant elements of biological complexity, such as social structure or mating behavior, as we do not have access to any relevant data to parameterize these, nor is any such information likely to be forthcoming.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Sex-associated PCFs were identified based on normalised depth using SATC, and all of them were confirmed to have been predicted to be part of the X chromosome.

    “We identified sex-associated PCFs based on the normalised depth of the samples mapped using SATC,101 and confirmed all of them had been predicted to be part of the X chromosome.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The positive outcome of genetic rescue in saola is contingent on the inferred demographic history characterized by a gradual population decline; a more sudden decline would have led to other conclusions.

    “We also demonstrate that the positive outcome of genetic rescue in saola is contingent on the inferred demographic history characterized by a gradual population decline and that a more sudden decline would have led to other conclusions.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Two complementary methods were used to estimate pairwise relatedness between samples.

    “We used two complementary methods to estimate pairwise relatedness between samples.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    Heterozygous calls within ROH regions are majoritarily due to genotyping errors or a minority of very recent de novo mutations that have no information on past coalescence rates.

    “We excluded any sites within these ROH where samples had an heterozygous call, that in a ROH region are majoritarily due to genotyping errors, or a minority of very recent de novo mutations that have no information on past coalescence rates.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    The recent decline in saola population sizes appears to be the continuation of a millennia-old decline, possibly instigated by a combination of climate change and early human activities.

    “The recent decline in saola population sizes hence appears to be the continuation of a millennia-old decline, possibly instigated by a combination of climate change and early human activities.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    Genotypes were called only in 8 samples for which the average depth after filtering was greater than 6X and the estimated error rate in transversion mutations was lower than 0.001.

    “We called genotypes in 8 samples for which the average depth after filtering was > 6X and the estimated error rate in transversion mutations was lower than 0.001.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    In small populations, genetic drift leads to an increase in mildly and slightly deleterious mutations, while purging reduces overall genetic load by removing strongly deleterious recessive mutations that contribute most to inbreeding depression.

    “In small populations, genetic drift leads to an increase in the mildly and slightly deleterious mutations, while purging reduces overall genetic load by removing the strongly deleterious recessive mutations that contribute the most to inbreeding depression.14”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [14] Dussex N. Morales H.E. Grossen C. Dalén L. van Oosterhout C. Purging and accumulation of genetic load in conservation Trends Ecol. Evol. 38 2023 961 969 10.1016/j.tree.2023.05.008 37344276 source

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    The results should solely be interpreted as a quantification of how genetic load and demographic stochasticity would progress in the population under the assumptions of the specific past demographic history and already accumulated genetic load.

    “Our results should therefore solely be interpreted as a quantification of how genetic load and demographic stochasticity (i.e., the possibility of fixation of one of the sexes) would progress in the population under the assumptions of the specific past demographic history and already accumulated genetic load.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationown data

    Assuming there are still some individuals surviving in the wild, the genomic dataset generated here should help improve new non-invasive DNA-based monitoring techniques such as screening for saola DNA from leech blood meals or using environmental DNA.

    “However, assuming there are still some individuals surviving in the wild, the genomic dataset generated here should help improve new non-invasive DNA-based monitoring techniques such as screening for saola DNA from leech blood meals80 or using environmental DNA.81”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [80] Schnell I.B. Thomsen P.F. Wilkinson N. Rasmussen M. Jensen L.R.D. Willerslev E. Bertelsen M.F. Gilbert M.T.P. Screening mammal biodiversity using DNA from leeches Curr. Biol. 22 2012 R262 R263 10.1016/j.cub.2012.02.058 22537625 source
    • cites [81] Thomsen P.F. Willerslev E. Environmental DNA – An emerging tool in conservation for monitoring past and present biodiversity Biol. Conserv. 183 2015 4 18 source
    • The authors state this as an assumption: There are still some individuals of saola surviving in the wild.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The bone samples were handled separately from the rest of the samples in the ancient DNA facilities at the Globe Institute, University of Copenhagen.

    “The bone samples were handled separately from the rest of samples in the ancient DNA facilities at the Globe Institute, University of Copenhagen.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    Heterozygosity estimates were considered reliable for samples with sequencing depth higher than 6 and transversion error rates lower than 0.001.

    “Based on visual inspection of this, we considered the heterozygosity estimates for samples with sequencing depth higher than 6 and transversion error rates in transversion mutations lower than 0.001.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The dried tissue samples were processed under strict clean laboratory conditions at the Globe Institute, University of Copenhagen.

    “The dried tissue samples were processed under strict clean laboratory conditions at the Globe Institute, University of Copenhagen.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    All genome-wide heterozygosity estimates in the compiled dataset quantify the same parameter: the proportion of nucleotide positions that are heterozygous for biallelic SNPs across a diploid genome.

    “All of these quantify the same parameter: the proportion of nucleotide positions that are heterozygous for biallelic SNPs across a diploid genome.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [130] Liu S. Westbury M.V. Dussex N. Mitchell K.J. Sinding M.S. Heintzman P.D. Duchêne D.A. Kapp J.D. von Seth J. Heiniger H. Ancient and modern genomes unravel the evolutionary history of the rhinoceros family Cell 184 2021 4874 4885.e16 10.1016/j.cell.2021.07.032 34433011 source

  • May combine two claims.

    Interpretationreasoning

    The observations of increased realized load and purging are not mutually contradictory and are consistent with recent findings regarding the typical outcome of reduced population sizes.

    “Rather than being mutually contradictory, these two observations are consistent with recent findings regarding the typical outcome of reduced population sizes.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning

    To mitigate the risk of errors or artifacts from sub-optimal DNA, the authors took extreme care in considering potential errors or artifacts in the sequencing data and only used a subset of samples that proved reliable and were sequenced to higher depth for the most sensitive analyses, e.g., ROH estimation and PSMC.

    “To mitigate this, we took extreme care in considering the potential for errors or artifacts in the sequencing data when conducting analyses and only used a subset of samples that proved reliable and were sequenced to higher depth for the analyses that are most sensitive, e.g., ROH estimation and PSMC.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    One principal component was detected as optimal by the built-in Velicer's minimum average partial test from PCAngsd for estimating population allele frequencies.

    “We used 1 principal component in the PCAngsd iterative algorithm to estimate population allele frequencies (-e 1), which was detected as optimal by the built-in Velicer's minimum average partial test from PCAngsd.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [103] Meisner J. Albrechtsen A. Inferring Population Structure and Admixture Proportions in Low-Depth NGS Data Genetics 210 2018 719 731 10.1534/genetics.118.301336 30131346 PMC6216594 source

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    There is outstanding uncertainty in which models better fit the relationship between selection coefficient s and dominance coefficient h observed in natural populations.

    “the outstanding uncertainty in which models better fit that observed in natural populations”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Using the folded 2dSFS minimizes potential bias arising when determining ancestral allelic states.

    “To minimize potential bias arising when determining ancestral allelic states, we used the folded 2dSFS, based on the whole genome 2DSFS estimated with winsfs as previously described.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationreasoning

    A relative decrease in genetic load indicates that there is purging.

    “A relative decrease in genetic load will indicate that there is purging.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Using as reference the generation before the decrease in population sizes allows measuring the relative change in genetic load due to demographic changes.

    “By using as reference the generation before the decrease in population sizes, we can measure the relative change in genetic load due to demographic changes.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    The marginal and joint distribution of selection coefficient s and dominance coefficient h can potentially influence the dynamics of genetic load.

    “Given the influence the marginal and joint distribution of s and h can potentially have in the dynamics of genetic load”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisereasoning

    Genetic load does not depend on demographic changes by themselves, as it is not affected by changes in genome-wide heterozygosity, but only depends on the number of derived alleles.

    “Note that the genetic load does not depend on demographic changes by themselves as it is not affected by changes in genome-wide heterozygosity, but only depends on the number of derived alleles.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisereasoning

    The two deleterious mutation models differ in the average deleteriousness of mutations and in the relationship between selection coefficient s and dominance coefficient h.

    “The models therefore differ in the average deleteriousness of mutations, with the Pérez-Pereira model assuming more deleterious mutations, and in the relationship between s and h.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Pérez-Pereira et al., 48

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Kyriazis et al., 146

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationreasoning

    If selection is not removing deleterious variants, genetic load will not be affected, while if there is genetic purging, genetic load will be reduced.

    “Therefore, if selection is not removing deleterious variants then the genetic load will not be affected, while if there is genetic purging then the genetic load will be reduced.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Both deleterious mutation models agree with the broadly observed negative relationship between dominance coefficient h and selection coefficient s, but the Pérez-Pereira model allows for more variable h such that less deleterious mutations can potentially be recessive or partially recessive, in opposition to the Kyriazis model.

    “While both agree with the broadly observed negative relationship between h and s, the Pérez-Pereira model allows for more variable h such that less deleterious mutations can potentially be recessive or partially recessive, in opposition to the Kyriazis model.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Pérez-Pereira et al., 48

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Kyriazis et al., 146

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

79 of 79

  • The paper relies on this without support.

    Before this study, the demographic history, population structure, genetic diversity, and genetic load of the saola were entirely unknown due to extremely limited genetic data.

    “However, due to the extremely limited genetic data from saola, these processes are entirely unknown in the species, which is a serious limitation for deciding appropriate conservation strategies.”

  • The authors state this as an assumption.

    There are still some individuals of saola surviving in the wild.

    “However, assuming there are still some individuals surviving in the wild, the genomic dataset generated here should help improve new non-invasive DNA-based monitoring techniques such as screening for saola DNA from leech blood meals80 or using environmental DNA.81”

  • The paper relies on this without support.

    The wet evergreen forest habitat in the ranges of the two saola populations is similar.

    “Therefore, and due to the similarity of the wet evergreen forest habitat in the ranges of the two populations, it is unlikely that the two populations have developed any significant local adaptations or genetic incompatibilities as a result of their divergence.”

  • The authors state this as an assumption.

    The saola reference genome consensus sequence used as the 'perfect individual' is error-free.

    “Error rates are estimated as excess distance of the samples to an outgroup, relative to the distance from a sample’s consensus sequence that is assumed to be error free and the outgroup. In this case, we used ‘perfect individual’ the saola reference genome, and used the cattle consensus sequence (see previous subsection) as the outgroup.”

  • The authors state this as an assumption.

    The mutation rate is 1.2e-8 mutations per generation.

    “We used a mutation rate of 1.2e-8 mutations/generation141 and a generation time of 6 years142 to scale the estimates of demographic history.”

  • The authors state this as an assumption.

    The generation time is 6 years.

    “We used a mutation rate of 1.2e-8 mutations/generation141 and a generation time of 6 years142 to scale the estimates of demographic history.”

  • The authors state this as an assumption.

    For SFS-based analyses where the SFS was estimated excluding transition mutations, the mutation rate is 4e-9 mutations per generation.

    “For SFS-based analyses where the SFS had been estimated excluding transition mutations, we used a rescaled mutation rate with the expected 1:3 ratio to 4e-9 mutations/generation.”

  • The authors state this as an assumption.

    The divergence model between the two populations has no gene flow.

    “We assumed a simple model only considering divergence time without gene flow between the two populations.”

  • The authors state this as an assumption.

    The simulation model assumes non-overlapping generations.

    “we used a simple model with non-overlapping generations”
9 of 9

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Population sample counts exceed the reported total

    Minor

    The listed geographic sample counts add to 27, but the paper says it retained 26 individuals. The subsequent discussion of a sample with uncertain origin does not reconcile these stated counts.

  • Opposite ratios are specified for relative cross-coalescence

    Minor

    The Figure S7 caption says to divide the between-population PSMC estimates by the within-population estimates. Those estimates are effective population sizes, which the caption says are inversely proportional to coalescence rates; the methods instead specify between-population coalescence rates divided by within-population rates.

  • Phylogenetic windows are described as both non-overlapping and overlapping

    Minor

    The results call the phylogenetic windows non-overlapping, but the methods specify 100-kb windows advanced in 50-kb steps, which overlap by 50 kb.

  • Two totals are given for the transversion SNP dataset

    Minor

    The results report 628,905 common transversion SNPs across 26 samples, whereas the SNP-calling methods report 644,546 transversion SNPs for the same 26-sample dataset. No distinction between these totals is stated.

4 of 4

Cites

Flagged first, then by citations

Cited by

Still being checked.

Related

Authors
Garcia-Erill, G., Liu, S., Le, M. D., Hurley, M. M., Nguyen, H. D., Nguyen, D. Q., Nguyen, D. H., Santander, C. G., Sánchez Barreiro, F., Gomes Martins, N. F., Hanghøj, K., Salleh, F. M., Ramos-Madrigal, J., Wang, X., Sinding, M.-H. S., Morales, H. E., Stæger, F. F., Wilkinson, N., Meng, G., Pečnerová, P., Yang, C., Rasmussen, M. S., Schubert, M., Dunn, R. R., Moltke, I., Zhang, G., Chen, L., Wang, W., Cao, T. T., Nguyen, H. M., Siegismund, H. R., Albrechtsen, A., Gilbert, M. T. P., & Heller, R.
Affiliations
  • University of Copenhagen (DK)
  • American Museum of Natural History (US)
  • Ministry of Agriculture and Rural Development (VN)
  • Zoological Research Museum Alexander Koenig (DE)
  • BGI Research (CN)
  • North Carolina State University (US)
  • Northwestern Polytechnical University (CN)
  • Vinh University (VN)
Corresponding author
  • Albrechtsen, Anders
  • Gilbert, M. Thomas P.
  • Heller, Rasmus
ORCID
Venue
Cell
Publisher
Elsevier BV
Published
2025-06
Volume
188
Issue
12
Pages
3102-3116.e22
ISSN
0092-8674 (ISSN-L)
DOI
10.1016/j.cell.2025.03.040
PMC
PMC12173715
OpenAlex
W4410091865
Cited by
10
References
139
Citations by year
2026: 8 · 2025: 2
FWCI
5.52
Type
journal article
Language
en
Open access
Licence
Locations
  • Cell (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • Research at the University of Copenhagen (University of Copenhagen) (publishedVersion, cc-by-nc-nd)
  • PubMed Central (submittedVersion, cc-by-nc-nd)
Funders
  • Independent Research Fund Denmark (DFF8049-00098B, DNRF143)
  • National Natural Science Foundation of China (32470455)
  • Novo Nordisk Foundation (NNF23SA0084103, NNF20OC0061343)
  • Danmarks Frie Forskningsfond
  • Institute of Zoology Chinese Academy of Sciences
  • European Research Council (681396)
  • Socialist Republic of Vietnam Ministry of Science and Technology (CN-64/19)
  • Danmarks Grundforskningsfond
  • Carlsberg Foundation (CF20-0355)
  • Chinese Academy of Sciences (2023IOZ0104)
  • Villum Foundation (VKR023447)
  • Disney Conservation Fund
  • Ministry of Science and Technology of the People's Republic of China (2023YFD2201804)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic diversity and population structure
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic and phenotypic traits in livestock
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
Keywords
critically endangered species, reference genome, population structure, genetic differentiation, purging, genetic load, Bovini, population divergence, climate change, anthropogenic impacts, captive breeding, conservation genetics
MeSH terms
Animals, Genetics, Population, Phylogeny, Genetic Variation, Genome, Endangered Species

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-10.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-10.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review