Cody Martin
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Cody Martin has a total of 76 co-authors. Here are the top co-authors by number of shared publications:
- Karthik Anantharaman
Department of Bacteriology, University of Wisconsin-Madison
12 shared publications - Zhichao Zhou
Institute for Advanced Study, Shenzhen University, Shenzhen, China
4 shared publications - Anthony Gitter
Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison, Madison, WI, USA
3 shared publications - Patricia Q. Tran
Department of Bacteriology, University of Wisconsin-Madison , Madison , WI , 53706
3 shared publications - Anna-Louise Reysenbach
Biology Department, Portland State University
2 shared publications - Ashley Holt
Department of Medicine, University of Wisconsin-Madison
2 shared publications
Latest Publications
- Open Access
CheckAMG: Accurate Identification of Auxiliary Viral Genes with Genome Language Models
Viruses shape microbial metabolism through auxiliary viral genes (AVGs) that reprogram host metabolism, physiology, and gene regulation during infection. Identifying AVGs is complicated by ambiguous viral versus cellular assignment, inconsistent curation, and reliance on sequence homology that misses divergent functions. Here we present CheckAMG, which pairs annotation-based AVG prediction, curated with a per-gene viral scoring system and ontological mapping of metabolic function, with a finetuned genome-language model for annotation-independent discovery. Benchmarked against DRAM-V and VIBRANT across three ecosystems, CheckAMG produced more strongly supported auxiliary metabolic gene (AMG) calls, and uniquely among AMG tools reports auxiliary physiological (APGs) and regulatory (AReGs) genes. Applied to 1,005,980 fragmented or complete soil and human-gut viral genomes, CheckAMG identified 541,199 AVGs tracking biome-specific protein clusters and functions, including 109,107 with no detectable sequence similarity to any reference database. CheckAMG thus provides a reproducible, scalable framework for identifying the genes through which viruses reprogram host biology.
Open AccessCheckAMG: Accurate Identification of Auxiliary Viral Genes with Genome Language Models
Viruses shape microbial metabolism through auxiliary viral genes (AVGs) that reprogram host metabolism, physiology, and gene regulation during infection. Identifying AVGs is complicated by ambiguous viral versus cellular assignment, inconsistent curation, and reliance on sequence homology that misses divergent functions. Here we present CheckAMG, which pairs annotation-based AVG prediction, curated with a per-gene viral scoring system and ontological mapping of metabolic function, with a finetuned genome-language model for annotation-independent discovery. Benchmarked against DRAM-V and VIBRANT across three ecosystems, CheckAMG produced more strongly supported auxiliary metabolic gene (AMG) calls, and uniquely among AMG tools reports auxiliary physiological (APGs) and regulatory (AReGs) genes. Applied to 1,005,980 fragmented or complete soil and human-gut viral genomes, CheckAMG identified 541,199 AVGs tracking biome-specific protein clusters and functions, including 109,107 with no detectable sequence similarity to any reference database. CheckAMG thus provides a reproducible, scalable framework for identifying the genes through which viruses reprogram host biology.
- Open Access
- 23 November 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1)
Protein Set Transformer: a protein-based genome language model to power high-diversity viromics
Exponential increases in microbial and viral genomic data demand transformational advances in scalable, generalizable frameworks for their interpretation. Standard homology-based functional analyses are hindered by the rapid divergence of microbial and especially viral genomes and proteins that significantly decreases the volume of usable data. Here, we present Protein Set Transformer (PST), a protein-based genome language model that models genomes as sets of proteins without considering sparsely available functional labels. Trained on >100k viruses, PST outperforms other homology- and language model-based approaches for relating viral genomes based on shared protein content. Further, PST demonstrates protein structural and functional awareness by clustering capsid-fold-containing proteins with known capsid proteins and uniquely clustering late gene proteins within related viruses. Our data establish PST as a valuable method for diverse viral genomics, ecology, and evolutionary applications. We posit that the PST framework can be a foundation model for microbial genomics when trained on suitable data.
- 23 November 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1)
Open AccessProtein Set Transformer: a protein-based genome language model to power high-diversity viromics
Exponential increases in microbial and viral genomic data demand transformational advances in scalable, generalizable frameworks for their interpretation. Standard homology-based functional analyses are hindered by the rapid divergence of microbial and especially viral genomes and proteins that significantly decreases the volume of usable data. Here, we present Protein Set Transformer (PST), a protein-based genome language model that models genomes as sets of proteins without considering sparsely available functional labels. Trained on >100k viruses, PST outperforms other homology- and language model-based approaches for relating viral genomes based on shared protein content. Further, PST demonstrates protein structural and functional awareness by clustering capsid-fold-containing proteins with known capsid proteins and uniquely clustering late gene proteins within related viruses. Our data establish PST as a valuable method for diverse viral genomics, ecology, and evolutionary applications. We posit that the PST framework can be a foundation model for microbial genomics when trained on suitable data.
- Open Access
Phage N4 uses a SAR endolysin-holin system for host cell lysis
Bacteriophages (phages) cause active host cell lysis to terminate their infection and release progeny into the environment. Some phages delay their lysis, and thereby increase progeny yield, in a phenomenon called lysis inhibition (LIN). There are two dissimilar phages of Escherichia coli known to exhibit LIN: T4 and N4. Studies of phage T4 have demonstrated a multi-protein mechanism that stalls lysis and maintains the LIN state in response to superinfection in a high phage population density. However, the lysis proteins responsible for T4 lysis and LIN are not conserved with phage N4. In this study, we characterize the phage N4 proteins involved in lysis by molecular and genetic means. We define the functions of the minimal gene set required for lysis through heterologous expression and complementation. Furthermore, by sequence comparison with a selected mutant library that does not induce LIN, we have identified genomic regions both within and outside the lysis cassette involved in N4 LIN. We propose a model where N4 lysis proteins that execute rapid N4 lysis can be regulated to induce LIN. Despite the lack of conservation with T4 components, our study suggests that direct modulation of lysis initiation may be common and provides a springboard for identifying other phages that regulate their phage production numbers in this way. Importance: Phages are viruses that kill only bacteria. To do so, they use lysis proteins targeted at the structural components of the bacterial envelope. There are commonalities between the phages that infect bacteria with similar structures, but there are also interesting differences we can exploit to learn more about specific phages or their bacterial hosts. In this study, we characterize the lysis proteins of the paradigm coliphage N4. Our results suggest that lysis regulation controls phage yield in N4 differently than in phage T4. Future studies on the relationship between lysis proteins and phage yield could provide handles to use this information for optimal large-scale phage production for clinical or industrial applications.
Open AccessPhage N4 uses a SAR endolysin-holin system for host cell lysis
Bacteriophages (phages) cause active host cell lysis to terminate their infection and release progeny into the environment. Some phages delay their lysis, and thereby increase progeny yield, in a phenomenon called lysis inhibition (LIN). There are two dissimilar phages of Escherichia coli known to exhibit LIN: T4 and N4. Studies of phage T4 have demonstrated a multi-protein mechanism that stalls lysis and maintains the LIN state in response to superinfection in a high phage population density. However, the lysis proteins responsible for T4 lysis and LIN are not conserved with phage N4. In this study, we characterize the phage N4 proteins involved in lysis by molecular and genetic means. We define the functions of the minimal gene set required for lysis through heterologous expression and complementation. Furthermore, by sequence comparison with a selected mutant library that does not induce LIN, we have identified genomic regions both within and outside the lysis cassette involved in N4 LIN. We propose a model where N4 lysis proteins that execute rapid N4 lysis can be regulated to induce LIN. Despite the lack of conservation with T4 components, our study suggests that direct modulation of lysis initiation may be common and provides a springboard for identifying other phages that regulate their phage production numbers in this way. Importance: Phages are viruses that kill only bacteria. To do so, they use lysis proteins targeted at the structural components of the bacterial envelope. There are commonalities between the phages that infect bacteria with similar structures, but there are also interesting differences we can exploit to learn more about specific phages or their bacterial hosts. In this study, we characterize the lysis proteins of the paradigm coliphage N4. Our results suggest that lysis regulation controls phage yield in N4 differently than in phage T4. Future studies on the relationship between lysis proteins and phage yield could provide handles to use this information for optimal large-scale phage production for clinical or industrial applications.
- Top Cited Papers
- 27 August 2025
- Published by Springer Nature in Nature Microbiology
- Vol. 10 (9) , 2122-2129
A call for caution in the biological interpretation of viral auxiliary metabolic genes
Virus-encoded auxiliary metabolic genes (AMGs) are non-essential genes that increase viral fitness by maintaining or manipulating host metabolism during infection. AMGs are intriguing from an evolutionary perspective, as most viral genomes are highly compact and have limited coding capacity for accessory genes. Advances in viral (meta)genomics have expanded the detection of putative AMGs from viruses in diverse environments. However, this has also led to many instances of misannotation due to the limitations of annotation tools, resulting in misinterpretations about the roles of some viral genes. Here, we highlight studies that support claims about AMGs with more than just function predictions for guidance on best practices. We then propose the adoption of an expanded, inclusive view of all genes auxiliary to core viral functions with the term ‘auxiliary viral genes’ (AVGs), alongside an associated eco-evolutionary framework for considering the types of analyses that can better support claims made about AVGs.
- 27 August 2025
- Published by Springer Nature in Nature Microbiology
- Vol. 10 (9) , 2122-2129
Top Cited PapersA call for caution in the biological interpretation of viral auxiliary metabolic genes
Virus-encoded auxiliary metabolic genes (AMGs) are non-essential genes that increase viral fitness by maintaining or manipulating host metabolism during infection. AMGs are intriguing from an evolutionary perspective, as most viral genomes are highly compact and have limited coding capacity for accessory genes. Advances in viral (meta)genomics have expanded the detection of putative AMGs from viruses in diverse environments. However, this has also led to many instances of misannotation due to the limitations of annotation tools, resulting in misinterpretations about the roles of some viral genes. Here, we highlight studies that support claims about AMGs with more than just function predictions for guidance on best practices. We then propose the adoption of an expanded, inclusive view of all genes auxiliary to core viral functions with the term ‘auxiliary viral genes’ (AVGs), alongside an associated eco-evolutionary framework for considering the types of analyses that can better support claims made about AVGs.
- Top Cited PapersOpen Access
- 1 May 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Endemism shapes viral ecology and evolution in globally distributed hydrothermal vent ecosystems
Viruses are ubiquitous in deep-sea hydrothermal vents, where they influence microbial communities and biogeochemistry. Yet, viral ecology and evolution remain understudied in these environments. Here, we identify 49,962 viruses from 52 globally distributed hydrothermal vent samples (10 plume, 40 deposit, and 2 diffuse flow metagenomes), and reconstruct 5708 viral metagenome-assembled genomes, the majority of which were bacteriophages. Hydrothermal viruses were largely endemic, however, some viruses were shared between geographically separated vents, predominantly between the Lau Basin and Brothers Volcano in the Pacific Ocean. Geographically distant viruses shared proteins related to core functions such as structural proteins, and rarely, proteins of auxiliary functions involved in processes such as fermentation and cobalamin biosynthesis. Common microbial hosts of viruses included members of Campylobacterota, Alpha-, and Gammaproteobacteria in deposits, and Gammaproteobacteria in plumes. Campylobacterota- and Gammaproteobacteria-infecting viruses reflected variations in hydrothermal chemistry and functional redundancy in their predicted microbial hosts, suggesting that hydrothermal geology is a driver of viral ecology and coevolution of viruses and hosts. Our results indicate that viral ecology and evolution in globally distributed hydrothermal vents is shaped by endemism and thus may have increased susceptibility to the negative impacts of deep-sea mining and anthropogenic change in ocean ecosystems.
- 1 May 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Top Cited PapersOpen AccessEndemism shapes viral ecology and evolution in globally distributed hydrothermal vent ecosystems
Viruses are ubiquitous in deep-sea hydrothermal vents, where they influence microbial communities and biogeochemistry. Yet, viral ecology and evolution remain understudied in these environments. Here, we identify 49,962 viruses from 52 globally distributed hydrothermal vent samples (10 plume, 40 deposit, and 2 diffuse flow metagenomes), and reconstruct 5708 viral metagenome-assembled genomes, the majority of which were bacteriophages. Hydrothermal viruses were largely endemic, however, some viruses were shared between geographically separated vents, predominantly between the Lau Basin and Brothers Volcano in the Pacific Ocean. Geographically distant viruses shared proteins related to core functions such as structural proteins, and rarely, proteins of auxiliary functions involved in processes such as fermentation and cobalamin biosynthesis. Common microbial hosts of viruses included members of Campylobacterota, Alpha-, and Gammaproteobacteria in deposits, and Gammaproteobacteria in plumes. Campylobacterota- and Gammaproteobacteria-infecting viruses reflected variations in hydrothermal chemistry and functional redundancy in their predicted microbial hosts, suggesting that hydrothermal geology is a driver of viral ecology and coevolution of viruses and hosts. Our results indicate that viral ecology and evolution in globally distributed hydrothermal vents is shaped by endemism and thus may have increased susceptibility to the negative impacts of deep-sea mining and anthropogenic change in ocean ecosystems.
- Top Cited Papers
- 3 January 2025
- Published by Springer Nature in Nature Microbiology
- Vol. 10 (1) , 231-245
Unravelling viral ecology and evolution over 20 years in a freshwater lake
As freshwater lakes undergo rapid anthropogenic change, long-term studies reveal key microbial dynamics, evolutionary shifts and biogeochemical interactions, yet the vital role of viruses remains overlooked. Here, leveraging a 20 year time series from Lake Mendota, WI, USA, we characterized 1.3 million viral genomes across time, seasonality and environmental factors. Double-stranded DNA phages from the class Caudoviricetes dominated the community. We identified 574 auxiliary metabolic gene families representing over 140,000 auxiliary metabolic genes, including important genes such as psbA (photosynthesis), pmoC (methane oxidation) and katG (hydrogen peroxide decomposition), which were consistently present and active across decades and seasons. Positive associations and niche differentiation between virus–host pairs, including keystone Cyanobacteria, methanotrophs and Nanopelagicales, emerged during seasonal changes. Inorganic carbon and ammonium influenced viral abundances, underscoring viral roles in both ‘top-down’ and ‘bottom-up’ interactions. Evolutionary processes favoured fitness genes, reduced genomic heterogeneity and dominant sub-populations. This study transforms understanding of viral ecology and evolution in Earth’s microbiomes.
- 3 January 2025
- Published by Springer Nature in Nature Microbiology
- Vol. 10 (1) , 231-245
Top Cited PapersUnravelling viral ecology and evolution over 20 years in a freshwater lake
As freshwater lakes undergo rapid anthropogenic change, long-term studies reveal key microbial dynamics, evolutionary shifts and biogeochemical interactions, yet the vital role of viruses remains overlooked. Here, leveraging a 20 year time series from Lake Mendota, WI, USA, we characterized 1.3 million viral genomes across time, seasonality and environmental factors. Double-stranded DNA phages from the class Caudoviricetes dominated the community. We identified 574 auxiliary metabolic gene families representing over 140,000 auxiliary metabolic genes, including important genes such as psbA (photosynthesis), pmoC (methane oxidation) and katG (hydrogen peroxide decomposition), which were consistently present and active across decades and seasons. Positive associations and niche differentiation between virus–host pairs, including keystone Cyanobacteria, methanotrophs and Nanopelagicales, emerged during seasonal changes. Inorganic carbon and ammonium influenced viral abundances, underscoring viral roles in both ‘top-down’ and ‘bottom-up’ interactions. Evolutionary processes favoured fitness genes, reduced genomic heterogeneity and dominant sub-populations. This study transforms understanding of viral ecology and evolution in Earth’s microbiomes.
- Open Access
- 23 September 2024
- Published by Springer Nature in Research Square
Protein Set Transformer: A protein-based genome language model to power high diversity viromics
Exponential increases in microbial and viral genomic data demand transformational advances in scalable, generalizable frameworks for their interpretation. Standard homology-based functional analyses are hindered by the rapid divergence of microbial and especially viral genomes and proteins that significantly decreases the volume of usable data. Here, we present Protein Set Transformer (PST), a protein-based genome language model that models genomes as sets of proteins without considering sparsely available functional labels. Trained on >100k viruses, PST outperformed other homology- and language model-based approaches for relating viral genomes based on shared protein content. Further, PST demonstrated protein structural and functional awareness by clustering capsid-fold-containing proteins with known capsid proteins and uniquely clustering late gene proteins within related viruses. Our data establish PST as a valuable method for diverse viral genomics, ecology, and evolutionary applications. We posit that the PST framework can be a foundation model for microbial genomics when trained on suitable data.
- 23 September 2024
- Published by Springer Nature in Research Square
Open AccessProtein Set Transformer: A protein-based genome language model to power high diversity viromics
Exponential increases in microbial and viral genomic data demand transformational advances in scalable, generalizable frameworks for their interpretation. Standard homology-based functional analyses are hindered by the rapid divergence of microbial and especially viral genomes and proteins that significantly decreases the volume of usable data. Here, we present Protein Set Transformer (PST), a protein-based genome language model that models genomes as sets of proteins without considering sparsely available functional labels. Trained on >100k viruses, PST outperformed other homology- and language model-based approaches for relating viral genomes based on shared protein content. Further, PST demonstrated protein structural and functional awareness by clustering capsid-fold-containing proteins with known capsid proteins and uniquely clustering late gene proteins within related viruses. Our data establish PST as a valuable method for diverse viral genomics, ecology, and evolutionary applications. We posit that the PST framework can be a foundation model for microbial genomics when trained on suitable data.
- Open Access
- 4 September 2024
- Published by Springer Nature in Research Square
Endemism shapes viral ecology and evolution in globally distributed hydrothermal vent ecosystems
Viruses are ubiquitous in deep-sea hydrothermal environments, where they exert a major influence on microbial communities and biogeochemistry. Yet, viral ecology and evolution remain understudied in these environments. Here, we identified 49,962 viruses from 52 globally distributed hydrothermal vent samples (10 plumes, 40 deposits, and 2 diffuse flow), and reconstructed 5,708 viral metagenome-assembled genomes (vMAGs), the majority of which were bacteriophages. Hydrothermal viruses were largely endemic. However, some viruses were shared between geographically separated vents, predominantly between the Lau Basin and Brothers Volcano in the Pacific Ocean. Geographically distant viruses often shared proteins related to core functions such as structural proteins, and rarely, proteins of auxiliary functions. Common microbial hosts of viruses included members of Campylobacterota, Alpha-, and Gammaproteobacteria in deposits, and Gammaproteobacteria in plumes. Campylobacterota- and Gammaproteobacteria-infecting viruses reflected variations in hydrothermal chemistry and functional redundancy in their predicted microbial hosts, suggesting that hydrothermal geology is a driver of viral ecology and coevolution of viruses and hosts. Our study indicates that viral ecology and evolution in globally distributed hydrothermal vents is shaped by endemism, and thus may have increased susceptibility to the negative impacts of deep-sea mining and anthropogenic change in ocean ecosystems.
- 4 September 2024
- Published by Springer Nature in Research Square
Open AccessEndemism shapes viral ecology and evolution in globally distributed hydrothermal vent ecosystems
Viruses are ubiquitous in deep-sea hydrothermal environments, where they exert a major influence on microbial communities and biogeochemistry. Yet, viral ecology and evolution remain understudied in these environments. Here, we identified 49,962 viruses from 52 globally distributed hydrothermal vent samples (10 plumes, 40 deposits, and 2 diffuse flow), and reconstructed 5,708 viral metagenome-assembled genomes (vMAGs), the majority of which were bacteriophages. Hydrothermal viruses were largely endemic. However, some viruses were shared between geographically separated vents, predominantly between the Lau Basin and Brothers Volcano in the Pacific Ocean. Geographically distant viruses often shared proteins related to core functions such as structural proteins, and rarely, proteins of auxiliary functions. Common microbial hosts of viruses included members of Campylobacterota, Alpha-, and Gammaproteobacteria in deposits, and Gammaproteobacteria in plumes. Campylobacterota- and Gammaproteobacteria-infecting viruses reflected variations in hydrothermal chemistry and functional redundancy in their predicted microbial hosts, suggesting that hydrothermal geology is a driver of viral ecology and coevolution of viruses and hosts. Our study indicates that viral ecology and evolution in globally distributed hydrothermal vents is shaped by endemism, and thus may have increased susceptibility to the negative impacts of deep-sea mining and anthropogenic change in ocean ecosystems.
- Open Access
Protein Set Transformer: A protein-based genome language model to power high diversity viromics
Exponential increases in microbial and viral genomic data demand transformational advances in scalable, generalizable frameworks for their interpretation. Standard homology-based functional analyses are hindered by the rapid divergence of microbial and especially viral genomes and proteins that significantly decreases the volume of usable data. Here, we present Protein Set Transformer (PST), a protein-based genome language model that models genomes as sets of proteins without considering sparsely available functional labels. Trained on >100k viruses, PST outperformed other homology- and language model-based approaches for relating viral genomes based on shared protein content. Further, PST demonstrated protein structural and functional awareness by clustering capsid-fold-containing proteins with known capsid proteins and uniquely clustering late gene proteins within related viruses. Our data establish PST as a valuable method for diverse viral genomics, ecology, and evolutionary applications. We posit that the PST framework can be a foundation model for microbial genomics when trained on suitable data.
Open AccessProtein Set Transformer: A protein-based genome language model to power high diversity viromics
Exponential increases in microbial and viral genomic data demand transformational advances in scalable, generalizable frameworks for their interpretation. Standard homology-based functional analyses are hindered by the rapid divergence of microbial and especially viral genomes and proteins that significantly decreases the volume of usable data. Here, we present Protein Set Transformer (PST), a protein-based genome language model that models genomes as sets of proteins without considering sparsely available functional labels. Trained on >100k viruses, PST outperformed other homology- and language model-based approaches for relating viral genomes based on shared protein content. Further, PST demonstrated protein structural and functional awareness by clustering capsid-fold-containing proteins with known capsid proteins and uniquely clustering late gene proteins within related viruses. Our data establish PST as a valuable method for diverse viral genomics, ecology, and evolutionary applications. We posit that the PST framework can be a foundation model for microbial genomics when trained on suitable data.
- Open Access
Systematic genome-wide discovery of host factors governing bacteriophage infectivity
Bacterial host factors regulate the infection cycle of bacteriophages. Except for some well-studied host factors (e.g., receptors or restriction-modification systems), the contribution of the rest of the host genome on phage infection remains poorly understood. We developed ‘PHAGEPACK’, a pooled assay that systematically and comprehensively measures each host gene’s impact on phage fitness. PHAGEPACK combines CRISPR interference with phage packaging to link host perturbation to phage fitness during active infection. Using PHAGEPACK, we constructed a genome-wide map of genes impacting T7 phage fitness in permissive E. coli , revealing pathways previously unknown to affect phage packaging. When applied to the non-permissive E. coli O121, PHAGEPACK identified pathways leading to host resistance; their removal increased phage susceptibility up to a billion-fold. Bioinformatic analysis indicates phage genomes carry homologs or truncations of key host factors, potentially for fitness advantage. In summary, PHAGEPACK offers valuable insights into phage-host interactions, phage evolution, and bacterial resistance.
Open AccessSystematic genome-wide discovery of host factors governing bacteriophage infectivity
Bacterial host factors regulate the infection cycle of bacteriophages. Except for some well-studied host factors (e.g., receptors or restriction-modification systems), the contribution of the rest of the host genome on phage infection remains poorly understood. We developed ‘PHAGEPACK’, a pooled assay that systematically and comprehensively measures each host gene’s impact on phage fitness. PHAGEPACK combines CRISPR interference with phage packaging to link host perturbation to phage fitness during active infection. Using PHAGEPACK, we constructed a genome-wide map of genes impacting T7 phage fitness in permissive E. coli , revealing pathways previously unknown to affect phage packaging. When applied to the non-permissive E. coli O121, PHAGEPACK identified pathways leading to host resistance; their removal increased phage susceptibility up to a billion-fold. Bioinformatic analysis indicates phage genomes carry homologs or truncations of key host factors, potentially for fitness advantage. In summary, PHAGEPACK offers valuable insights into phage-host interactions, phage evolution, and bacterial resistance.