James J. Davis

GenSLMs use large language models to analyze SARS-CoV-2 genomes, enabling rapid identification of variants of concern and providing insights into evolutionary dynamics.

This study builds machine learning classifiers to predict variable gene presence in E. coli genomes, achieving high accuracy for most gene families, including poorly annotated and horizontally transferred genes.

This paper develops a random forest model to predict plasmid hosts, achieving 0.662 and 0.867 Matthews correlation coefficients at species and order levels, respectively.

The paper compares ARTIC V3 and V4 primer sets for SARS-CoV-2 sequencing, finding improved performance and underrepresentation of G142D substitution in early Delta genomes due to V3 primer limitations.

The paper sequences 320 SARS-CoV-2 genomes from Houston to study early pandemic spread, finding diverse origins and no evidence of more virulent strains.

This study shows that machine learning can predict antimicrobial resistance using conserved non-AMR genes from incomplete genomes, achieving F1 scores of 0.80-0.89 with 100 genes.

This paper uses machine learning to predict the virulence of Pseudomonas aeruginosa isolates based on genomic data, achieving 75% accuracy using random forest models.

This study uses XGBoost to predict antimicrobial MICs for nontyphoidal Salmonella with high accuracy using genomic data.

Open-source KBase platform unites reference genomes and biochemistry, over 70 apps, Jupyter-based Narratives with provenance, and a Docker SDK for reproducible plant and microbial systems biology.

This study develops an XGBoost-based machine learning model to predict minimum inhibitory concentrations for 20 antibiotics in Klebsiella pneumoniae using whole genome sequencing data, achieving 92% overall accuracy.

The paper introduces a pan-synteny graph method for analyzing multiple prokaryotic genomes, enabling the identification of genomic changes and conservation patterns without a reference genome.