Merve N Güler, Ardan Yılmaz, Büşra Katırcıoğlu, Sarp Kantar, Tara Ekin Ünver, Kıvılcım Başak Vural, N Ezgi Altınışık, Emre Akbas, Mehmet Somel
{"title":"DeepKin:用卷积神经网络预测低覆盖率基因组和古基因组的相关性。","authors":"Merve N Güler, Ardan Yılmaz, Büşra Katırcıoğlu, Sarp Kantar, Tara Ekin Ünver, Kıvılcım Başak Vural, N Ezgi Altınışık, Emre Akbas, Mehmet Somel","doi":"10.1111/1755-0998.70032","DOIUrl":null,"url":null,"abstract":"<p><p>DeepKin is a novel tool designed to predict relatedness from genomic data using convolutional neural networks (CNNs). Traditional methods for estimating relatedness often struggle when genomic data is limited, as with palaeogenomes and degraded forensic samples. DeepKin addresses this challenge by leveraging two CNN models, which are trained solely on simulated genomic data, to classify relatedness up to the third degree and to identify parent-offspring and sibling pairs. Our benchmarking shows DeepKin performs comparably or better than the widely used tool READv2. We validated DeepKin, which uses PLINK's .map and .ped files as input, on empirical palaeogenomes from three archaeological sites, demonstrating its robustness and adaptability across different genetic backgrounds, with accuracy > 90% above 10 K shared SNPs. By capturing information across genomic segments, DeepKin offers a new methodological path for relatedness estimation in settings with highly degraded samples, with applications in ancient DNA, as well as forensic and conservation genetics.</p>","PeriodicalId":211,"journal":{"name":"Molecular Ecology Resources","volume":" ","pages":"e70032"},"PeriodicalIF":5.5000,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"DeepKin: Predicting Relatedness From Low-Coverage Genomes and Palaeogenomes With Convolutional Neural Networks.\",\"authors\":\"Merve N Güler, Ardan Yılmaz, Büşra Katırcıoğlu, Sarp Kantar, Tara Ekin Ünver, Kıvılcım Başak Vural, N Ezgi Altınışık, Emre Akbas, Mehmet Somel\",\"doi\":\"10.1111/1755-0998.70032\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><p>DeepKin is a novel tool designed to predict relatedness from genomic data using convolutional neural networks (CNNs). Traditional methods for estimating relatedness often struggle when genomic data is limited, as with palaeogenomes and degraded forensic samples. DeepKin addresses this challenge by leveraging two CNN models, which are trained solely on simulated genomic data, to classify relatedness up to the third degree and to identify parent-offspring and sibling pairs. Our benchmarking shows DeepKin performs comparably or better than the widely used tool READv2. We validated DeepKin, which uses PLINK's .map and .ped files as input, on empirical palaeogenomes from three archaeological sites, demonstrating its robustness and adaptability across different genetic backgrounds, with accuracy > 90% above 10 K shared SNPs. By capturing information across genomic segments, DeepKin offers a new methodological path for relatedness estimation in settings with highly degraded samples, with applications in ancient DNA, as well as forensic and conservation genetics.</p>\",\"PeriodicalId\":211,\"journal\":{\"name\":\"Molecular Ecology Resources\",\"volume\":\" \",\"pages\":\"e70032\"},\"PeriodicalIF\":5.5000,\"publicationDate\":\"2025-08-18\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Molecular Ecology Resources\",\"FirstCategoryId\":\"99\",\"ListUrlMain\":\"https://doi.org/10.1111/1755-0998.70032\",\"RegionNum\":1,\"RegionCategory\":\"生物学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"BIOCHEMISTRY & MOLECULAR BIOLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Molecular Ecology Resources","FirstCategoryId":"99","ListUrlMain":"https://doi.org/10.1111/1755-0998.70032","RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"BIOCHEMISTRY & MOLECULAR BIOLOGY","Score":null,"Total":0}
DeepKin: Predicting Relatedness From Low-Coverage Genomes and Palaeogenomes With Convolutional Neural Networks.
DeepKin is a novel tool designed to predict relatedness from genomic data using convolutional neural networks (CNNs). Traditional methods for estimating relatedness often struggle when genomic data is limited, as with palaeogenomes and degraded forensic samples. DeepKin addresses this challenge by leveraging two CNN models, which are trained solely on simulated genomic data, to classify relatedness up to the third degree and to identify parent-offspring and sibling pairs. Our benchmarking shows DeepKin performs comparably or better than the widely used tool READv2. We validated DeepKin, which uses PLINK's .map and .ped files as input, on empirical palaeogenomes from three archaeological sites, demonstrating its robustness and adaptability across different genetic backgrounds, with accuracy > 90% above 10 K shared SNPs. By capturing information across genomic segments, DeepKin offers a new methodological path for relatedness estimation in settings with highly degraded samples, with applications in ancient DNA, as well as forensic and conservation genetics.
期刊介绍:
Molecular Ecology Resources promotes the creation of comprehensive resources for the scientific community, encompassing computer programs, statistical and molecular advancements, and a diverse array of molecular tools. Serving as a conduit for disseminating these resources, the journal targets a broad audience of researchers in the fields of evolution, ecology, and conservation. Articles in Molecular Ecology Resources are crafted to support investigations tackling significant questions within these disciplines.
In addition to original resource articles, Molecular Ecology Resources features Reviews, Opinions, and Comments relevant to the field. The journal also periodically releases Special Issues focusing on resource development within specific areas.