Tracing the Enterococci from Paleozoic Origins to the Hospital.

We examined the evolutionary history of leading multidrug resistant hospital pathogens, the enterococci, to their origin hundreds of millions of years ago. Our goal was to understand why, among the vast diversity of gut flora, enterococci are so well adapted to the modern hospital environment. Molecular clock estimation, together with analysis of their environmental distribution, phenotypic diversity, and concordance with host fossil records, place the origins of the enterococci around the time of animal terrestrialization, 425-500 mya. Speciation appears to parallel the diversification of hosts, including the rapid emergence of new enterococcal species following the End Permian Extinction. Major drivers of speciation include changing carbohydrate availability in the host gut. Life on land would have selected for the precise traits that now allow pathogenic enterococci to survive desiccation, starvation, and disinfection in the modern hospital, foreordaining their emergence as leading hospital pathogens.

Discovery of a structural class of antibiotics with explainable deep learning.

The discovery of novel structural classes of antibiotics is urgently needed to address the ongoing antibiotic resistance crisis1-9. Deep learning approaches have aided in exploring chemical spaces1,10-15; these typically use black box models and do not provide chemical insights. Here we reasoned that the chemical substructures associated with antibiotic activity learned by neural network models can be identified and used to predict structural classes of antibiotics. We tested this hypothesis by developing an explainable, substructure-based approach for the efficient, deep learning-guided exploration of chemical spaces. We determined the antibiotic activities and human cell cytotoxicity profiles of 39,312 compounds and applied ensembles of graph neural networks to predict antibiotic activity and cytotoxicity for 12,076,365 compounds. Using explainable graph algorithms, we identified substructure-based rationales for compounds with high predicted antibiotic activity and low predicted cytotoxicity. We empirically tested 283 compounds and found that compounds exhibiting antibiotic activity against Staphylococcus aureus were enriched in putative structural classes arising from rationales. Of these structural classes of compounds, one is selective against methicillin-resistant S. aureus (MRSA) and vancomycin-resistant enterococci, evades substantial resistance, and reduces bacterial titres in mouse models of MRSA skin and systemic thigh infection. Our approach enables the deep learning-guided discovery of structural classes of antibiotics and demonstrates that machine learning models in drug discovery can be explainable, providing insights into the chemical substructures that underlie selective antibiotic activity.

Global diversity of enterococci and description of 18 previously unknown species.

Enterococci are gut microbes of most land animals. Likely appearing first in the guts of arthropods as they moved onto land, they diversified over hundreds of millions of years adapting to evolving hosts and host diets. Over 60 enterococcal species are now known. Two species, Enterococcus faecalis and Enterococcus faecium, are common constituents of the human microbiome. They are also now leading causes of multidrug-resistant hospital-associated infection. The basis for host association of enterococcal species is unknown. To begin identifying traits that drive host association, we collected 886 enterococcal strains from widely diverse hosts, ecologies, and geographies. This identified 18 previously undescribed species expanding genus diversity by >25%. These species harbor diverse genes including toxins and systems for detoxification and resource acquisition. Enterococcus faecalis and E. faecium were isolated from diverse hosts highlighting their generalist properties. Most other species showed a more restricted distribution indicative of specialized host association. The expanded species diversity permitted the Enterococcus genus phylogeny to be viewed with unprecedented resolution, allowing features to be identified that distinguish its four deeply rooted clades, and the entry of genes associated with range expansion such as B-vitamin biosynthesis and flagellar motility to be mapped to the phylogeny. This work provides an unprecedentedly broad and deep view of the genus Enterococcus, including insights into its evolution, potential new threats to human health, and where substantial additional enterococcal diversity is likely to be found.

Comparison of long- and short-read metagenomic assembly for low-abundance species and resistance genes.

Recent technological and computational advances have made metagenomic assembly a viable approach to achieving high-resolution views of complex microbial communities. In previous benchmarking, short-read (SR) metagenomic assemblers had the highest accuracy, long-read (LR) assemblers generated the most contiguous sequences and hybrid (HY) assemblers balanced length and accuracy. However, no assessments have specifically compared the performance of these assemblers on low-abundance species, which include clinically relevant organisms in the gut. We generated semi-synthetic LR and SR datasets by spiking small and increasing amounts of Escherichia coli isolate reads into fecal metagenomes and, using different assemblers, examined E. coli contigs and the presence of antibiotic resistance genes (ARGs). For ARG assembly, although SR assemblers recovered more ARGs with high accuracy, even at low coverages, LR assemblies allowed for the placement of ARGs within longer, E. coli-specific contigs, thus pinpointing their taxonomic origin. HY assemblies identified resistance genes with high accuracy and had lower contiguity than LR assemblies. Each assembler type's strengths were maintained even when our isolate was spiked in with a competing strain, which fragmented and reduced the accuracy of all assemblies. For strain characterization and determining gene context, LR assembly is optimal, while for base-accurate gene identification, SR assemblers outperform other options. HY assembly offers contiguity and base accuracy, but requires generating data on multiple platforms, and may suffer high misassembly rates when strain diversity exists. Our results highlight the trade-offs associated with each approach for recovering low-abundance taxa, and that the optimal approach is goal-dependent.

SAFPred: synteny-aware gene function prediction for bacteria using protein embeddings.

MOTIVATION: Today, we know the function of only a small fraction of the protein sequences predicted from genomic data. This problem is even more salient for bacteria, which represent some of the most phylogenetically and metabolically diverse taxa on Earth. This low rate of bacterial gene annotation is compounded by the fact that most function prediction algorithms have focused on eukaryotes, and conventional annotation approaches rely on the presence of similar sequences in existing databases. However, often there are no such sequences for novel bacterial proteins. Thus, we need improved gene function prediction methods tailored for bacteria. Recently, transformer-based language models-adopted from the natural language processing field-have been used to obtain new representations of proteins, to replace amino acid sequences. These representations, referred to as protein embeddings, have shown promise for improving annotation of eukaryotes, but there have been only limited applications on bacterial genomes. RESULTS: To predict gene functions in bacteria, we developed SAFPred, a novel synteny-aware gene function prediction tool based on protein embeddings from state-of-the-art protein language models. SAFpred also leverages the unique operon structure of bacteria through conserved synteny. SAFPred outperformed both conventional sequence-based annotation methods and state-of-the-art methods on multiple bacterial species, including for distant homolog detection, where the sequence similarity to the proteins in the training set was as low as 40%. Using SAFPred to identify gene functions across diverse enterococci, of which some species are major clinical threats, we identified 11 previously unrecognized putative novel toxins, with potential significance to human and animal health. AVAILABILITY AND IMPLEMENTATION: https://github.com/AbeelLab/safpred.

Utility of Mycobacterium tuberculosis genome sequencing snapshots to assess transmission dynamics over time.

We explored the utility of brief Mycobacterium tuberculosis whole-genome sequencing (WGS) "snapshots" at a sentinel site within Lima, Peru for evaluating local transmission dynamics over time. Within a 17 km2 area, 15/70 (21%) isolates with WGS collected during 2011-2012 and 22/81 (27%) collected during 2020-2021 were clustered (p = 0.414), and additional isolates clustered with those from outside the area. Isolates from the later period were disproportionately related to large historic clusters in Lima from the earlier period. WGS snapshots at a sentinel site may not be useful for monitoring transmission, but monitoring the persistence of large transmission clusters might be.

Benchmarking AlphaFold-enabled molecular docking predictions for antibiotic discovery.

Efficient identification of drug mechanisms of action remains a challenge. Computational docking approaches have been widely used to predict drug binding targets; yet, such approaches depend on existing protein structures, and accurate structural predictions have only recently become available from AlphaFold2. Here, we combine AlphaFold2 with molecular docking simulations to predict protein-ligand interactions between 296 proteins spanning Escherichia coli's essential proteome, and 218 active antibacterial compounds and 100 inactive compounds, respectively, pointing to widespread compound and protein promiscuity. We benchmark model performance by measuring enzymatic activity for 12 essential proteins treated with each antibacterial compound. We confirm extensive promiscuity, but find that the average area under the receiver operating characteristic curve (auROC) is 0.48, indicating weak model performance. We demonstrate that rescoring of docking poses using machine learning-based approaches improves model performance, resulting in average auROCs as large as 0.63, and that ensembles of rescoring functions improve prediction accuracy and the ratio of true-positive rate to false-positive rate. This work indicates that advances in modeling protein-ligand interactions, particularly using machine learning-based approaches, are needed to better harness AlphaFold2 for drug discovery.

Limited effects of long-term daily cranberry consumption on the gut microbiome in a placebo-controlled study of women with recurrent urinary tract infections.

BACKGROUND: Urinary tract infections (UTIs) affect 15 million women each year in the United States, with > 20% experiencing frequent recurrent UTIs. A recent placebo-controlled clinical trial found a 39% reduction in UTI symptoms among recurrent UTI sufferers who consumed a daily cranberry beverage for 24 weeks. Using metagenomic sequencing of stool from a subset of these trial participants, we assessed the impact of cranberry consumption on the gut microbiota, a reservoir for UTI-causing pathogens such as Escherichia coli, which causes > 80% of UTIs. RESULTS: The overall taxonomic composition, community diversity, carriage of functional pathways and gene families, and relative abundances of the vast majority of observed bacterial taxa, including E. coli, were not changed significantly by cranberry consumption. However, one unnamed Flavonifractor species (OTU41), which represented ≤1% of the overall metagenome, was significantly less abundant in cranberry consumers compared to placebo at trial completion. Given Flavonifractor's association with negative human health effects, we sought to determine OTU41 characteristic genes that may explain its differential abundance and/or relationship to key host functions. Using comparative genomic and metagenomic techniques, we identified genes in OTU41 related to transport and metabolism of various compounds, including tryptophan and cobalamin, which have been shown to play roles in host-microbe interactions. CONCLUSION: While our results indicated that cranberry juice consumption had little impact on global measures of the microbiome, we found one unnamed Flavonifractor species differed significantly between study arms. This suggests further studies are needed to assess the role of cranberry consumption and Flavonifractor in health and wellbeing in the context of recurrent UTI. TRIAL REGISTRATION: Clinical trial registration number: ClinicalTrials.gov NCT01776021 .

QuantTB - a method to classify mixed Mycobacterium tuberculosis infections within whole genome sequencing data.

BACKGROUND: Mixed infections of Mycobacterium tuberculosis and antibiotic heteroresistance continue to complicate tuberculosis (TB) diagnosis and treatment. Detection of mixed infections has been limited to molecular genotyping techniques, which lack the sensitivity and resolution to accurately estimate the multiplicity of TB infections. In contrast, whole genome sequencing offers sensitive views of the genetic differences between strains of M. tuberculosis within a sample. Although metagenomic tools exist to classify strains in a metagenomic sample, most tools have been developed for more divergent species, and therefore cannot provide the sensitivity required to disentangle strains within closely related bacterial species such as M. tuberculosis. Here we present QuantTB, a method to identify and quantify individual M. tuberculosis strains in whole genome sequencing data. QuantTB uses SNP markers to determine the combination of strains that best explain the allelic variation observed in a sample. QuantTB outputs a list of identified strains, their corresponding relative abundances, and a list of drugs for which resistance-conferring mutations (or heteroresistance) have been predicted within the sample. RESULTS: We show that QuantTB has a high degree of resolution and is capable of differentiating communities differing by less than 25 SNPs and identifying strains down to 1× coverage. Using simulated data, we found QuantTB outperformed other metagenomic strain identification tools at detecting strains and quantifying strain multiplicity. In a real-world scenario, using a dataset of 50 paired clinical isolates from a study of patients with either reinfections or relapses, we found that QuantTB could detect mixed infections and reinfections at rates concordant with a manually curated approach. CONCLUSION: QuantTB can determine infection multiplicity, identify hetero-resistance patterns, enable differentiation between relapse and re-infection, and clarify transmission events across seemingly unrelated patients - even in low-coverage (1×) samples. QuantTB outperforms existing tools and promises to serve as a valuable resource for both clinicians and researchers working with clinical TB samples.

Evidence for Expanding the Role of Streptomycin in the Management of Drug-Resistant Mycobacterium tuberculosis.

In 2019, the WHO tuberculosis (TB) treatment guidelines were updated to recommend only limited use of streptomycin, in favor of newer agents or amikacin as the preferred aminoglycoside for drug-resistant Mycobacterium tuberculosis However, the emergence of resistance to newer drugs, such as bedaquiline, has prompted a reanalysis of antitubercular drugs in search of untapped potential. Using 211 clinical isolates of M. tuberculosis from South Africa, we performed phenotypic drug susceptibility testing (DST) to aminoglycosides by both critical concentration and MIC determination in parallel with whole-genome sequencing to identify known genotypic resistance elements. Isolates with low-level streptomycin resistance mediated by gidB were frequently misclassified with respect to streptomycin resistance when using the WHO-recommended critical concentration of 2 µg/ml. We identified 29 M. tuberculosis isolates from South Africa with low-level streptomycin resistance concomitant with high-level amikacin resistance, conferred by gidB and rrs 1400, respectively. Using a large global data set of M. tuberculosis genomes, we observed 95 examples of this corresponding resistance genotype (gidB-rrs 1400), including identification in 81/257 (31.5%) of extensively drug resistant (XDR) isolates. In a phylogenetic analysis, we observed repeated evolution of low-level streptomycin and high-level amikacin resistance in multiple countries. Our findings suggest that current critical concentration methods and the design of molecular diagnostics need to be revisited to provide more accurate assessments of streptomycin resistance for gidB-containing isolates. For patients harboring isolates of M. tuberculosis with high-level amikacin resistance conferred by rrs 1400, and for whom newer agents are not available, treatment with streptomycin may still prove useful, even in the face of low-level resistance conferred by gidB.

Extensive global movement of multidrug-resistantM. tuberculosis strains revealed by whole-genome analysis.

BACKGROUND: While the international spread of multidrug-resistant (MDR) Mycobacterium tuberculosis strains is an acknowledged public health threat, a broad and more comprehensive examination of the global spread of MDR-tuberculosis (TB) using whole-genome sequencing has not yet been performed. METHODS: In a global dataset of 5310 M. tuberculosis whole-genome sequences isolated from five continents, we performed a phylogenetic analysis to identify and characterise clades of MDR-TB with respect to geographic dispersion. RESULTS: Extensive international dissemination of MDR-TB was observed, with identification of 32 migrant MDR-TB clades with descendants isolated in 17 unique countries. Relatively recent movement of strains from both Beijing and non-Beijing lineages indicated successful global spread of varied genetic backgrounds. Migrant MDR-TB clade members shared relatively recent common ancestry, with a median estimate of divergence of 13-27 years. Migrant extensively drug-resistant (XDR)-TB clades were not observed, although development of XDR-TB within migratory MDR-TB clades was common. CONCLUSIONS: Application of genomic techniques to investigate global MDR migration patterns revealed extensive global spread of MDR clades between countries of varying TB burden. Further expansion of genomic studies to incorporate isolates from diverse global settings into a single analysis, as well as data sharing platforms that facilitate genomic data sharing across country lines, may allow for future epidemiological analyses to monitor for international transmission of MDR-TB. In addition, efforts to perform routine whole-genome sequencing on all newly identified M. tuberculosis, like in England, will serve to better our understanding of the transmission dynamics of MDR-TB globally.

Chicken Meat-Associated Enterococci: Influence of Agricultural Antibiotic Use and Connection to the Clinic.

Industrial farms are unique, human-created ecosystems that provide the perfect setting for the development and dissemination of antibiotic resistance. Agricultural antibiotic use amplifies naturally occurring resistance mechanisms from soil ecologies, promoting their spread and sharing with other bacteria, including those poised to become endemic within hospital environments. To better understand the role of enterococci in the movement of antibiotic resistance from farm to table to clinic, we characterized over 300 isolates of Enterococcus cultured from raw chicken meat purchased at U.S. supermarkets by the Consumers Union in 2013. Enterococcus faecalis and Enterococcus faecium were the predominant species found, and antimicrobial susceptibility testing uncovered striking levels of resistance to medically important antibiotic classes, particularly from classes approved by the FDA for use in animal production. While nearly all isolates were resistant to at least one drug, bacteria from meat labeled as raised without antibiotics had fewer resistances, particularly for E. faecium Whole-genome sequencing of 92 isolates revealed that both commensal- and clinical-isolate-like enterococcal strains were associated with chicken meat, including isolates bearing important resistance-conferring elements and virulence factors. The ability of enterococci to persist in the food system positions them as vehicles to move resistance genes from the industrial farm ecosystem into more human-proximal ecologies.IMPORTANCE Bacteria that contaminate food can serve as a conduit for moving drug resistance genes from farm to table to clinic. Our results show that chicken meat-associated isolates of Enterococcus are often multidrug resistant, closely related to pathogenic lineages, and harbor worrisome virulence factors. These drug-resistant agricultural isolates could thus represent important stepping stones in the evolution of enterococci into drug-resistant human pathogens. Although significant efforts have been made over the past few years to reduce the agricultural use of antibiotics, continued assessment of agricultural practices, including the roles of processing plants, shared breeding flocks, and probiotics as sources for resistance spread, is needed in order to slow the evolution of antibiotic resistance. Because antibiotic resistance is a global problem, global policies are needed to address this threat. Additional measures must be taken to mitigate the development and spread of antibiotic resistance elements from farms to clinics throughout the world.

Mycobacterium tuberculosis Whole Genome Sequences From Southern India Suggest Novel Resistance Mechanisms and the Need for Region-Specific Diagnostics.

BACKGROUND.: India is home to 25% of all tuberculosis cases and the second highest number of multidrug resistant cases worldwide. However, little is known about the genetic diversity and resistance determinants of Indian Mycobacterium tuberculosis, particularly for the primary lineages found in India, lineages 1 and 3. METHODS.: We whole genome sequenced 223 randomly selected M. tuberculosis strains from 196 patients within the Tiruvallur and Madurai districts of Tamil Nadu in Southern India. Using comparative genomics, we examined genetic diversity, transmission patterns, and evolution of resistance. RESULTS.: Genomic analyses revealed (11) prevalence of strains from lineages 1 and 3, (11) recent transmission of strains among patients from the same treatment centers, (11) emergence of drug resistance within patients over time, (11) resistance gained in an order typical of strains from different lineages and geographies, (11) underperformance of known resistance-conferring mutations to explain phenotypic resistance in Indian strains relative to studies focused on other geographies, and (11) the possibility that resistance arose through mutations not previously implicated in resistance, or through infections with multiple strains that confound genotype-based prediction of resistance. CONCLUSIONS.: In addition to substantially expanding the genomic perspectives of lineages 1 and 3, sequencing and analysis of M. tuberculosis whole genomes from Southern India highlight challenges of infection control and rapid diagnosis of resistant tuberculosis using current technologies. Further studies are needed to fully explore the complement of diversity and resistance determinants within endemic M. tuberculosis populations.

Whole-Transcriptome and -Genome Analysis of Extensively Drug-Resistant Mycobacterium tuberculosis Clinical Isolates Identifies Downregulation of ethA as a Mechanism of Ethionamide Resistance.

Genetics-based drug susceptibility testing has improved the diagnosis of drug-resistant tuberculosis but is limited by our lack of knowledge of all resistance mechanisms. Next-generation sequencing has assisted in identifying the principal genetic mechanisms of resistance for many drugs, but a significant proportion of phenotypic drug resistance is unexplained genetically. Few studies have formally compared the transcriptomes of susceptible and resistant Mycobacterium tuberculosis strains. We carried out comparative whole-genome transcriptomics of extensively drug-resistant (XDR) clinical isolates using RNA sequencing (RNA-seq) to find novel transcription-mediated mechanisms of resistance. We identified a promoter mutation (t to c) at position -11 (t-11c) relative to the start codon of ethA that reduces the expression of a monooxygenase (EthA) that activates ethionamide. (In this article, nucleotide changes are lowercase and amino acid substitutions are uppercase.) Using a flow cytometry-based reporter assay, we show that the reduced transcription of ethA is not due to transcriptional repression by ethR Clinical strains harboring this mutation were resistant to ethionamide. Other ethA promoter mutations were identified in a global genomic survey of resistant M. tuberculosis strains. These results demonstrate a new mechanism of ethionamide resistance that can cause high-level resistance when it is combined with other ethionamide resistance-conferring mutations. Our study revealed many other genes which were highly up- or downregulated in XDR strains, including a toxin-antitoxin module (mazF5 mazE5) and tRNAs (leuX and thrU). This suggests that global transcriptional modifications could contribute to resistance or the maintenance of bacterial fitness have also occurred in XDR strains.

Persistent Salmonella infections in humans are associated with mutations in the BarA/SirA regulatory pathway.

Several bacterial pathogens, including Salmonella enterica, can cause persistent infections in humans by mechanisms that are poorly understood. By comparing genomes of isolates longitudinally collected from 256 prolonged salmonellosis patients, we identified repeated mutations in global regulators, including the barA/sirA two-component regulatory system, across multiple patients and Salmonella serovars. Comparative RNA-seq analysis revealed that distinct mutations in barA/sirA led to diminished expression of Salmonella pathogenicity islands 1 and 4 genes, which are required for Salmonella invasion and enteritis. Moreover, barA/sirA mutants were attenuated in an acute salmonellosis mouse model and induced weaker transcription of host immune responses. In contrast, in a persistent infection mouse model, these mutants exhibited long-term colonization and prolonged shedding. Taken together, these findings suggest that selection of mutations in global virulence regulators facilitates persistent Salmonella infection in humans, by attenuating Salmonella virulence and inducing a weaker host inflammatory response.

Distinct Escherichia coli transcriptional profiles in the guts of recurrent UTI sufferers revealed by pangenome hybrid selection.

Low-abundance members of microbial communities are difficult to study in their native habitats. This includes Escherichia coli, a minor, but common inhabitant of the gastrointestinal tract and opportunistic pathogen, including of the urinary tract, where it is the primary pathogen. While multi-omic analyses have detailed critical interactions between uropathogenic Escherichia coli (UPEC) and the bladder that mediate UTI outcome, comparatively little is known about UPEC in its pre-infection reservoir, partly due to its low abundance there (<1% relative abundance). To accurately and sensitively explore the genomes and transcriptomes of diverse E. coli in gastrointestinal communities, we developed E. coli PanSelect which uses a set of probes designed to specifically recognize and capture E. coli's broad pangenome from sequencing libraries. We demonstrated the ability of E. coli PanSelect to enrich, by orders of magnitude, sequencing data from diverse E. coli using a mock community and a set of human stool samples collected as part of a cohort study investigating drivers of recurrent urinary tract infections (rUTI). Comparisons of genomes and transcriptomes between E. coli residing in the gastrointestinal tracts of women with and without a history of rUTI suggest that rUTI gut E. coli are responding to increased levels of oxygen and nitrate, suggestive of mucosal inflammation, which may have implications for recurrent disease. E. coli PanSelect is well suited for investigations of native in vivo biology of E. coli in other environments where it is at low relative abundance, and the framework described here has broad applicability to other highly diverse, low abundance organisms.

Genomic epidemiology and antibiotic susceptibility profiling of uropathogenic Escherichia coli among children in the United States.

Escherichia coli is the most common cause of urinary tract infections (UTIs) in children, and yet the underlying mechanisms of virulence and antibiotic resistance and the overall population structure of the species is poorly understood within this age group. To investigate whether uropathogenic E. coli (UPEC) from children who developed pyelonephritis carried specific genetic markers, we generated whole-genome sequence data for 96 isolates from children with UTIs. This included 57 isolates from children with either radiologically confirmed pyelonephritis or cystitis and 27 isolates belonging to the well-known multidrug-resistant sequence type ST131, selected to investigate their population structure and antibiotic resistance characteristics. We observed a UPEC population structure that is similar to those reported in adults. In comparison with prior investigations, we found that the full pap operon was more common among UPEC from pediatric cases of pyelonephritis. Further, in contrast with recent reports that the P-fimbriae adhesin-encoding papGII allele is substantially more prevalent in invasive UPEC from adults, we found papGII was common to both invasive and non-invasive UPEC from children. Among the set of ST131 isolates from children with UTIs, we found antibiotic resistance was correlated with known genetic markers of resistance, as in adults. Unexpectedly, we observed that fimH30, an allele of the fimbrial gene fimH often used as a proxy to type ST131 isolates into the most drug-resistant subclade C, was carried by some of the subclade A and subclade B isolates, suggesting that the fimH30 allele could confer a selective advantage for UPEC. IMPORTANCE Urinary tract infections (UTIs), which are most often caused by Escherichia coli, are not well studied in children. Here, we examine genetic characteristics that differentiate UTI-causing bacteria in children that either remain localized to the bladder or are involved in more serious kidney infections. We also examine patterns of antibiotic resistance among strains from children that are part of E. coli sequence type 131, a group of bacteria that commonly cause UTIs and are known to have high levels of drug resistance. This work provides new insight into the virulence and antibiotic resistance characteristics of the bacteria that cause UTIs in children.

SAP: Synteny-aware gene function prediction for bacteria using protein embeddings.

Motivation: Today, we know the function of only a small fraction of the protein sequences predicted from genomic data. This problem is even more salient for bacteria, which represent some of the most phylogenetically and metabolically diverse taxa on Earth. This low rate of bacterial gene annotation is compounded by the fact that most function prediction algorithms have focused on eukaryotes, and conventional annotation approaches rely on the presence of similar sequences in existing databases. However, often there are no such sequences for novel bacterial proteins. Thus, we need improved gene function prediction methods tailored for prokaryotes. Recently, transformer-based language models - adopted from the natural language processing field - have been used to obtain new representations of proteins, to replace amino acid sequences. These representations, referred to as protein embeddings, have shown promise for improving annotation of eukaryotes, but there have been only limited applications on bacterial genomes. Results: To predict gene functions in bacteria, we developed SAP, a novel synteny-aware gene function prediction tool based on protein embeddings from state-of-the-art protein language models. SAP also leverages the unique operon structure of bacteria through conserved synteny. SAP outperformed both conventional sequence-based annotation methods and state-of-the-art methods on multiple bacterial species, including for distant homolog detection, where the sequence similarity to the proteins in the training set was as low as 40%. Using SAP to identify gene functions across diverse enterococci, of which some species are major clinical threats, we identified 11 previously unrecognized putative novel toxins, with potential significance to human and animal health.

zol & fai: large-scale targeted detection and evolutionary investigation of gene clusters.

Many universally and conditionally important genes are genomically aggregated within clusters. Here, we introduce fai and zol, which together enable large-scale comparative analysis of different types of gene clusters and mobile-genetic elements (MGEs), such as biosynthetic gene clusters (BGCs) or viruses. Fundamentally, they overcome a current bottleneck to reliably perform comprehensive orthology inference at large scale across broad taxonomic contexts and thousands of genomes. First, fai allows the identification of orthologous or homologous instances of a query gene cluster of interest amongst a database of target genomes. Subsequently, zol enables reliable, context-specific inference of protein-encoding ortholog groups for individual genes across gene cluster instances. In addition, zol performs functional annotation and computes a variety of statistics for each inferred ortholog group. These programs are showcased through application to: (i) longitudinal tracking of a virus in metagenomes, (ii) discovering novel population-genetic insights of two common BGCs in a fungal species, and (iii) uncovering large-scale evolutionary trends of a virulence-associated gene cluster across thousands of genomes from a diverse bacterial genus.