Jacob Lucas Steenwyk

The paper analyzes how ESM3, a multimodal protein language model, organizes different modalities across its depth, finding that physical modalities fuse into a shared subspace while functional annotations remain orthogonal.

The paper shows that structure tokens in ESM-3 enhance the shared biological vocabulary of protein language models without creating a new feature set.

The paper isolates Candida auris from Antarctica, identifies cold-adapted traits, and links mutator phenotypes to rapid evolution and drug resistance in global C. auris clades.

This study shows that the choanoflagellate Choanoeca flexa forms multicellular sheets through clonal or aggregative mechanisms, adapting to environmental changes in splash pools.

PAQman evaluates long-read eukaryotic genome assemblies using seven reference-free metrics without requiring a reference genome.

The paper discovers three novel ancient whole genome duplications in Saccharomycotina yeasts, suggesting WGD may be more common in fungi than previously thought.

This study compares the genomes of nematode-trapping fungi and non-NTF to identify genomic changes linked to carnivorism and adaptation to post-Permian-Triassic extinction environments.

This study uses AlphaFold2 to analyze how metabolism influenced enzyme structural evolution across 400 million years by examining 11,269 enzyme structures and their metabolic reactions.

The paper analyzes gene family evolution across nearly all known Saccharomycotina yeast species, finding distinct evolutionary patterns with gene loss linked to metabolic and genomic changes.

The paper identifies three yeast lineages with reduced DNA repair genes and high evolutionary rates, showing how gene loss can lead to persistent hypermutator phenotypes.

OrthoHMM improves ortholog group inference using Hidden Markov Models, showing significant precision gains over existing methods in benchmark tests.

RCVT quantifies compositional biases among taxa in phylogenomic data, showing that removing high RCVT taxa reduces compositional heterogeneity.

The paper uses phylogenomic analyses of 23 genomes to study evolutionary modes of carnivorous fungi, finding extensive phylogenetic discordance due to incomplete lineage sorting and non-vertical evolution.

Solu provides a cloud-based platform for real-time genomic pathogen surveillance with comparable accuracy to existing pipelines and identifies previously undetected antimicrobial resistance genes.

CryptoCEN, a co-expression network for Cryptococcus neoformans, identifies 13 novel DNA damage repair genes with 100% predictive accuracy.

The paper introduces LVBRS, a cloud-based toolkit for bulk RNA-seq analysis, demonstrating its utility in identifying differentially expressed genes and functional insights in hypoxia models.

The paper examines genome-wide ortholog variation in 206 A. fumigatus isolates, finding no significant differences in ortholog distribution between clinical and environmental isolates after accounting for phylogeny.

BioKIT provides a unified toolkit for processing and analyzing diverse sequence data, demonstrating its utility through quality assessments and codon usage analysis across multiple datasets.

The paper identifies AlkZ and AlkX DNA glycosylases involved in bacterial resistance to genotoxins, linking them to specific antimicrobial production.

integRATE prioritizes candidate genes across heterogeneous omics data by using desirability functions, identifying known and novel genes in preterm birth.

The paper constructs a robust phylogenomic timetree for Aspergillus and Penicillium fungi using multiple genome datasets and phylogenetic methods to identify incongruent branches and estimate divergence times.

The paper presents a whole genome DNA methylation profile of Dictyostelium discoideum, revealing a sparse but robust pattern of 5-methyl-cytosine associated with genes and retrotransposable elements.