Felix M. Key
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Felix M. Key has a total of 482 co-authors. Here are the top co-authors by number of shared publications:
- Alexander Herbig
Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany
10 shared publications - Johannes Krause
Department of Archaeogenetics, Max Planck Institute for Evolutionary Anthropology, Leipzig 04103, Germany
10 shared publications - Aida M. Andres
UCL Genetics Institute, Department of Genetics, Evolution and Environment, University College London, London, UK
9 shared publications - Ian Light-Maka
Evolutionary Pathogenomics, Max Planck Institute for Infection Biology, Berlin, Germany
9 shared publications - Kirsten I. Bos
Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany
7 shared publications - Christina Warinner
Department of Anthropology, Harvard University, Cambridge, MA 02138, USA
5 shared publications
Latest Publications
- Open Access
- 4 September 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 12 (36)
Three thousand five hundred years of sheeppox virus evolution inferred from archaeological and codicological genomes
Sheeppox virus (SPPV) is a major livestock pathogen causing economic hardship through reduced production and death of vulnerable sheep, with written descriptions of sheeppox-like disease recorded since antiquity. We report 21 novel ancient SPPV genomes spanning the Eurasian steppe Bronze Age (∼1700 BCE) to the Early Modern period in Western Europe, including multiple genomes obtained from medieval parchment. We estimate that major capripoxvirus lineages diverged ∼11,500 to 3700 years ago, overlapping known translocations and bio-cultural developments in sheep. Our dataset supports SPPV diverging first within the lineage leading to goatpox virus and lumpy skin disease virus, and that known gene inactivation events within SPPV and goatpox virus occur in our earliest SPPV genomes. These findings reveal that the food security of Eurasian communities has been threatened by sheeppox for more than 3700 years and provide insights into the genomic evolution and potential host adaptation of SPPV.
- 4 September 2026
- Published by American Association for the Advancement of Science (AAAS) in Science Advances
- Vol. 12 (36)
Open AccessThree thousand five hundred years of sheeppox virus evolution inferred from archaeological and codicological genomes
Sheeppox virus (SPPV) is a major livestock pathogen causing economic hardship through reduced production and death of vulnerable sheep, with written descriptions of sheeppox-like disease recorded since antiquity. We report 21 novel ancient SPPV genomes spanning the Eurasian steppe Bronze Age (∼1700 BCE) to the Early Modern period in Western Europe, including multiple genomes obtained from medieval parchment. We estimate that major capripoxvirus lineages diverged ∼11,500 to 3700 years ago, overlapping known translocations and bio-cultural developments in sheep. Our dataset supports SPPV diverging first within the lineage leading to goatpox virus and lumpy skin disease virus, and that known gene inactivation events within SPPV and goatpox virus occur in our earliest SPPV genomes. These findings reveal that the food security of Eurasian communities has been threatened by sheeppox for more than 3700 years and provide insights into the genomic evolution and potential host adaptation of SPPV.
- 6 July 2026
- Published by Cold Spring Harbor Laboratory in Genome Research
- Vol. 36 (9) , 1889-1901
High-accuracy SNV calling for bacterial isolates using deep learning with AccuSNV
Accurate detection of mutations within bacterial species is critical for fundamental studies of microbial evolution, reconstruction of transmission events, and identification of antimicrobial resistance mutations. Although many tools have been developed to identify single-nucleotide variants (SNVs) from whole-genome sequencing, they often suffer from high false-positive rates owing to the complexity of bacterial genomes and the need for different filtering cutoffs across sample types and sequencing depths. As data sets increase in size, the manual filtering required for high accuracy presents a significant obstacle. Here, we present AccuSNV, a novel deep learning–based tool for high-precision and automated bacterial SNV calling. Unlike traditional methods that process one sample at a time, AccuSNV leverages a convolutional neural network (CNN) that integrates alignment information across multiple samples, enhancing precision through learned across-sample patterns. We evaluate AccuSNV against seven popular SNV-calling tools using simulated data from six bacterial species with varied sequencing depths, numbers of isolates, mutations, and divergence levels. To further validate its real-world utility, we test AccuSNV on multiple curated bacterial data sets containing reported SNVs. In both simulated and real-world scenarios, AccuSNV consistently achieves the best performance. Moreover, AccuSNV provides comprehensive user-friendly downstream analysis modules and outputs, including mutation annotation information, phylogenetic inference, d N / d S calculations, and optional manual filtering. Together with the automated deep learning–based calling, these features make AccuSNV broadly accessible to users with different levels of computational expertise.
- 6 July 2026
- Published by Cold Spring Harbor Laboratory in Genome Research
- Vol. 36 (9) , 1889-1901
High-accuracy SNV calling for bacterial isolates using deep learning with AccuSNV
Accurate detection of mutations within bacterial species is critical for fundamental studies of microbial evolution, reconstruction of transmission events, and identification of antimicrobial resistance mutations. Although many tools have been developed to identify single-nucleotide variants (SNVs) from whole-genome sequencing, they often suffer from high false-positive rates owing to the complexity of bacterial genomes and the need for different filtering cutoffs across sample types and sequencing depths. As data sets increase in size, the manual filtering required for high accuracy presents a significant obstacle. Here, we present AccuSNV, a novel deep learning–based tool for high-precision and automated bacterial SNV calling. Unlike traditional methods that process one sample at a time, AccuSNV leverages a convolutional neural network (CNN) that integrates alignment information across multiple samples, enhancing precision through learned across-sample patterns. We evaluate AccuSNV against seven popular SNV-calling tools using simulated data from six bacterial species with varied sequencing depths, numbers of isolates, mutations, and divergence levels. To further validate its real-world utility, we test AccuSNV on multiple curated bacterial data sets containing reported SNVs. In both simulated and real-world scenarios, AccuSNV consistently achieves the best performance. Moreover, AccuSNV provides comprehensive user-friendly downstream analysis modules and outputs, including mutation annotation information, phylogenetic inference, d N / d S calculations, and optional manual filtering. Together with the automated deep learning–based calling, these features make AccuSNV broadly accessible to users with different levels of computational expertise.
- Open Access
- 30 April 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Probing the zooarchaeological record across time and space for ancient pathogen DNA
Zoonoses are among the greatest threats to human health, with many zoonotic pathogens believed to have emerged following the Neolithic transition. Palaeomicrobiological investigations of the zooarchaeological record hold potential to uncover the reservoirs, host ranges, and host adaptations of zoonotic pathogens in the past, but face challenges in identifying promising specimens and pathogen DNA preservation. We perform palaeopathological and genetic examinations of 346 skeletal elements from domesticated and wild animals collected from 34 Eurasian sites dating across the last six millennia. We identify 116 signatures of 29 ancient (opportunistic) pathogens and find support that palaeopathological lesions provide guidance for specimen selection. For two pathogen species, Erysipelothrix rhusiopathiae and Streptococcus lutetiensis , we confirm their ancient authenticity using phylogenetics, showcasing an approach to explore the relationship between ancient low-coverage genomes and their modern-day relatives. Our work presents a pathway to understanding prehistoric zoonotic diseases by integrating zooarchaeological, palaeopathological, and genetic data.
- 30 April 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessProbing the zooarchaeological record across time and space for ancient pathogen DNA
Zoonoses are among the greatest threats to human health, with many zoonotic pathogens believed to have emerged following the Neolithic transition. Palaeomicrobiological investigations of the zooarchaeological record hold potential to uncover the reservoirs, host ranges, and host adaptations of zoonotic pathogens in the past, but face challenges in identifying promising specimens and pathogen DNA preservation. We perform palaeopathological and genetic examinations of 346 skeletal elements from domesticated and wild animals collected from 34 Eurasian sites dating across the last six millennia. We identify 116 signatures of 29 ancient (opportunistic) pathogens and find support that palaeopathological lesions provide guidance for specimen selection. For two pathogen species, Erysipelothrix rhusiopathiae and Streptococcus lutetiensis , we confirm their ancient authenticity using phylogenetics, showcasing an approach to explore the relationship between ancient low-coverage genomes and their modern-day relatives. Our work presents a pathway to understanding prehistoric zoonotic diseases by integrating zooarchaeological, palaeopathological, and genetic data.
- Open Access
3,500 years of sheeppox virus evolution inferred from archaeological and codicological genomes
Sheeppox virus (SPPV) is a major livestock pathogen causing economic hardship through reduced production and death of vulnerable sheep, with written descriptions of sheeppox-like disease recorded since antiquity. We report 21 novel ancient SPPV genomes spanning the Eurasian steppe Bronze Age (∼1,700 BCE) to the Early Modern period in Western Europe, including multiple genomes obtained from medieval parchment. We estimate that major capripoxvirus lineages diverged ∼11,500-3,700 years ago, overlapping known translocations and bio-cultural developments in sheep. Our dataset supports SPPV diverging first within the lineage leading to goatpox virus and lumpy skin disease virus, and that known gene inactivation events within SPPV and goatpox virus occur in our earliest SPPV genomes. These findings reveal that the food security of Eurasian communities have been threatened by sheeppox for over 3,700 years, and provide new insights to the genomic evolution and potential host adaptation of sheeppox virus. Teaser: Ancient viral genomes from manuscripts and teeth illuminate the evolutionary history of sheeppox virus.
Open Access3,500 years of sheeppox virus evolution inferred from archaeological and codicological genomes
Sheeppox virus (SPPV) is a major livestock pathogen causing economic hardship through reduced production and death of vulnerable sheep, with written descriptions of sheeppox-like disease recorded since antiquity. We report 21 novel ancient SPPV genomes spanning the Eurasian steppe Bronze Age (∼1,700 BCE) to the Early Modern period in Western Europe, including multiple genomes obtained from medieval parchment. We estimate that major capripoxvirus lineages diverged ∼11,500-3,700 years ago, overlapping known translocations and bio-cultural developments in sheep. Our dataset supports SPPV diverging first within the lineage leading to goatpox virus and lumpy skin disease virus, and that known gene inactivation events within SPPV and goatpox virus occur in our earliest SPPV genomes. These findings reveal that the food security of Eurasian communities have been threatened by sheeppox for over 3,700 years, and provide new insights to the genomic evolution and potential host adaptation of sheeppox virus. Teaser: Ancient viral genomes from manuscripts and teeth illuminate the evolutionary history of sheeppox virus.
- Open Access
Colonization, translocation, and evolution of opportunistic pathogens during healthcare-associated infections
Many commensal bacteria that peacefully reside in the human microbiome are also able to cause acute opportunistic infections. Emerging evidence suggests that within-host evolution contributes to infection, but the genetic mechanisms facilitating the progression of opportunistic pathogens from carriage to acute infection remain unknown. Here, we prospectively collected native samples from four microbiome niches of 13 critically ill patients to assess the evolutionary dynamics leading up to infection. Among three patients we have observed eleven healthcare-associated infections (HAI) caused by nine pathogen species. Leveraging a culture-based approach, we demonstrate that the microbiome is frequently (73%) colonized by the pathogen lineage already before or at the time of diagnosis. Moreover, we identify a short-lived, non-synonymous mutation (F126L) within the fimbriae regulator gene fimZ of Enterobacter hormaechei , first detectable within the gut and subsequently associated with HAI before becoming replaced by body-wide sweeps of independent treatment-associated mutations. Despite fimZ [F126L] being globally undetected, we can show in vitro and in vivo that the F126L mutation leads to elevated biofilm formation, cell adhesion and virulence, suggesting a role during HAI. Our work highlights the power of prospective, population-wide investigation of pathogens to elucidate rapid evolution linked to disease.
Open AccessColonization, translocation, and evolution of opportunistic pathogens during healthcare-associated infections
Many commensal bacteria that peacefully reside in the human microbiome are also able to cause acute opportunistic infections. Emerging evidence suggests that within-host evolution contributes to infection, but the genetic mechanisms facilitating the progression of opportunistic pathogens from carriage to acute infection remain unknown. Here, we prospectively collected native samples from four microbiome niches of 13 critically ill patients to assess the evolutionary dynamics leading up to infection. Among three patients we have observed eleven healthcare-associated infections (HAI) caused by nine pathogen species. Leveraging a culture-based approach, we demonstrate that the microbiome is frequently (73%) colonized by the pathogen lineage already before or at the time of diagnosis. Moreover, we identify a short-lived, non-synonymous mutation (F126L) within the fimbriae regulator gene fimZ of Enterobacter hormaechei , first detectable within the gut and subsequently associated with HAI before becoming replaced by body-wide sweeps of independent treatment-associated mutations. Despite fimZ [F126L] being globally undetected, we can show in vitro and in vivo that the F126L mutation leads to elevated biofilm formation, cell adhesion and virulence, suggesting a role during HAI. Our work highlights the power of prospective, population-wide investigation of pathogens to elucidate rapid evolution linked to disease.
- Open Access
High-accuracy SNV calling for bacterial isolates using deep learning with AccuSNV
Accurate detection of mutations within bacterial species is critical for fundamental studies of microbial evolution, reconstructing transmission events, and identifying antimicrobial resistance mutations. While many tools have been developed to identify single nucleotide variants (SNVs) from whole-genome sequencing, they often suffer from high false positive rates due to the complexity of bacterial genomes and the need for different filtering cutoffs across sample types and sequencing depths. As datasets increase in size, the manual filtering required for high accuracy presents a significant obstacle. Here, we present AccuSNV, a novel deep learning-based tool for high-precision and automated bacterial SNV calling. Unlike traditional methods that process one sample at a time, AccuSNV leverages a convolutional neural network (CNN) that integrates alignment information across multiple samples, enhancing precision through learned across-sample patterns. We evaluated AccuSNV against seven popular SNV calling tools using simulated data from six bacterial species with varied sequencing depths, numbers of isolates, mutations, and divergence levels. To further validate its real-world utility, we tested AccuSNV on multiple curated bacterial datasets containing reported SNVs. In both simulated and real-world scenarios, AccuSNV consistently achieved the best performance. Moreover, AccuSNV provides comprehensive user-friendly downstream analysis modules and outputs, including mutation annotation information, phylogenetic inference, dN/dS calculations, and optional manual filtering. Together with the automated deep learning–based calling, these features make AccuSNV broadly accessible to users with different levels of computational expertise.
Open AccessHigh-accuracy SNV calling for bacterial isolates using deep learning with AccuSNV
Accurate detection of mutations within bacterial species is critical for fundamental studies of microbial evolution, reconstructing transmission events, and identifying antimicrobial resistance mutations. While many tools have been developed to identify single nucleotide variants (SNVs) from whole-genome sequencing, they often suffer from high false positive rates due to the complexity of bacterial genomes and the need for different filtering cutoffs across sample types and sequencing depths. As datasets increase in size, the manual filtering required for high accuracy presents a significant obstacle. Here, we present AccuSNV, a novel deep learning-based tool for high-precision and automated bacterial SNV calling. Unlike traditional methods that process one sample at a time, AccuSNV leverages a convolutional neural network (CNN) that integrates alignment information across multiple samples, enhancing precision through learned across-sample patterns. We evaluated AccuSNV against seven popular SNV calling tools using simulated data from six bacterial species with varied sequencing depths, numbers of isolates, mutations, and divergence levels. To further validate its real-world utility, we tested AccuSNV on multiple curated bacterial datasets containing reported SNVs. In both simulated and real-world scenarios, AccuSNV consistently achieved the best performance. Moreover, AccuSNV provides comprehensive user-friendly downstream analysis modules and outputs, including mutation annotation information, phylogenetic inference, dN/dS calculations, and optional manual filtering. Together with the automated deep learning–based calling, these features make AccuSNV broadly accessible to users with different levels of computational expertise.
- Open Access
Probing the zooarchaeological record across time and space for ancient pathogens
Zoonoses are among the greatest threats to human health, with many zoonotic pathogens believed to have emerged during prehistory. Palaeomicrobiological investigations of the zooarchaeological record hold potential to uncover the reservoirs, host ranges, and host adaptations of zoonotic pathogens but face challenges in identifying promising specimens and pathogen DNA preservation. We performed palaeopathological and genetic examinations of 346 skeletal elements from domesticated and wild animals collected from 34 Eurasian sites dating across the last six millennia. We identified 68 signatures of ancient (opportunistic) pathogens, including the important zoonotic pathogen Salmonella enterica , and found support that palaeopathological lesions provide guidance for specimen selection. For two pathogen species, Erysipelothrix rhusiopathiae and Streptococcus lutetiensis , we confirmed their ancient authenticity using phylogenetics, showcasing an approach to explore the relationship between ancient low-coverage genomes and their modern-day relatives. Our work presents a pathway to understanding prehistoric zoonotic diseases by integrating zooarchaeological, palaeopathological, and genetic data.
Open AccessProbing the zooarchaeological record across time and space for ancient pathogens
Zoonoses are among the greatest threats to human health, with many zoonotic pathogens believed to have emerged during prehistory. Palaeomicrobiological investigations of the zooarchaeological record hold potential to uncover the reservoirs, host ranges, and host adaptations of zoonotic pathogens but face challenges in identifying promising specimens and pathogen DNA preservation. We performed palaeopathological and genetic examinations of 346 skeletal elements from domesticated and wild animals collected from 34 Eurasian sites dating across the last six millennia. We identified 68 signatures of ancient (opportunistic) pathogens, including the important zoonotic pathogen Salmonella enterica , and found support that palaeopathological lesions provide guidance for specimen selection. For two pathogen species, Erysipelothrix rhusiopathiae and Streptococcus lutetiensis , we confirmed their ancient authenticity using phylogenetics, showcasing an approach to explore the relationship between ancient low-coverage genomes and their modern-day relatives. Our work presents a pathway to understanding prehistoric zoonotic diseases by integrating zooarchaeological, palaeopathological, and genetic data.
- Open Access
Bronze Age Yersinia pestis genome from sheep sheds light on hosts and evolution of a prehistoric plague lineage
Summary: Most human pathogens are of zoonotic origin. Many emerged during prehistory, coinciding with domestication providing more opportunities for spillover from original host species. However, we lack direct evidence linking past animal reservoirs and human infections. Here we present a Yersinia pestis genome recovered from a 3rd millennium BCE domesticated sheep from the Eurasian Steppe belonging to the Late Neolithic Bronze Age (LNBA) lineage, until now exclusively identified in ancient humans across Eurasia. We show that this ancient lineage underwent ancestral gene decay paralleling extant lineages, but evolved under distinct selective pressures contributing to its lack of geographic differentiation. We collect evidence supporting a scenario where the LNBA lineage, unable to efficiently transmit via fleas, spread from an unidentified reservoir to humans via sheep and likely other domesticates. Collectively, our results connect prehistoric livestock with infectious disease in humans and showcase the power of moving paleomicrobiology into the zooarcheological record.
Open AccessBronze Age Yersinia pestis genome from sheep sheds light on hosts and evolution of a prehistoric plague lineage
Summary: Most human pathogens are of zoonotic origin. Many emerged during prehistory, coinciding with domestication providing more opportunities for spillover from original host species. However, we lack direct evidence linking past animal reservoirs and human infections. Here we present a Yersinia pestis genome recovered from a 3rd millennium BCE domesticated sheep from the Eurasian Steppe belonging to the Late Neolithic Bronze Age (LNBA) lineage, until now exclusively identified in ancient humans across Eurasia. We show that this ancient lineage underwent ancestral gene decay paralleling extant lineages, but evolved under distinct selective pressures contributing to its lack of geographic differentiation. We collect evidence supporting a scenario where the LNBA lineage, unable to efficiently transmit via fleas, spread from an unidentified reservoir to humans via sheep and likely other domesticates. Collectively, our results connect prehistoric livestock with infectious disease in humans and showcase the power of moving paleomicrobiology into the zooarcheological record.
- Open Access
- 20 July 2024
- Published by Springer Nature in Nature Communications
- Vol. 15 (1) , 1-10
An 8000 years old genome reveals the Neolithic origin of the zoonosis Brucella melitensis
Brucella melitensis is a major livestock bacterial pathogen and zoonosis, causing disease and infection-related abortions in small ruminants and humans. A considerable burden to animal-based economies today, the presence of Brucella in Neolithic pastoral communities has been hypothesised but we lack direct genomic evidence thus far. We report a 3.45X B. melitensis genome preserved in an ~8000 year old sheep specimen from Menteşe Höyük, Northwest Türkiye, demonstrating that the pathogen had evolved and was circulating in Neolithic livestock. The genome is basal with respect to all known B. melitensis and allows the calibration of the B. melitensis speciation time from the primarily cattle-infecting B. abortus to approximately 9800 years Before Present (BP), coinciding with a period of consolidation and dispersal of livestock economies. We use the basal genome to timestamp evolutionary events in B. melitensis, including pseudogenization events linked to erythritol response, the supposed determinant of the pathogen’s placental tropism in goats and sheep. Our data suggest that the development of herd management and multi-species livestock economies in the 11th–9th millennium BP drove speciation and host adaptation of this zoonotic pathogen.
- 20 July 2024
- Published by Springer Nature in Nature Communications
- Vol. 15 (1) , 1-10
Open AccessAn 8000 years old genome reveals the Neolithic origin of the zoonosis Brucella melitensis
Brucella melitensis is a major livestock bacterial pathogen and zoonosis, causing disease and infection-related abortions in small ruminants and humans. A considerable burden to animal-based economies today, the presence of Brucella in Neolithic pastoral communities has been hypothesised but we lack direct genomic evidence thus far. We report a 3.45X B. melitensis genome preserved in an ~8000 year old sheep specimen from Menteşe Höyük, Northwest Türkiye, demonstrating that the pathogen had evolved and was circulating in Neolithic livestock. The genome is basal with respect to all known B. melitensis and allows the calibration of the B. melitensis speciation time from the primarily cattle-infecting B. abortus to approximately 9800 years Before Present (BP), coinciding with a period of consolidation and dispersal of livestock economies. We use the basal genome to timestamp evolutionary events in B. melitensis, including pseudogenization events linked to erythritol response, the supposed determinant of the pathogen’s placental tropism in goats and sheep. Our data suggest that the development of herd management and multi-species livestock economies in the 11th–9th millennium BP drove speciation and host adaptation of this zoonotic pathogen.