Phillip Davis

The paper develops a model using ORF1ab codon frequencies to predict host-pathogen relationships in Orthocoronavirinae viruses with high precision and recall.

The paper develops a genotype-to-phenotype model to predict the human pathogenicity of novel coronaviruses using nucleotide k-mer features and shows it can accurately classify known pathogens and non-pathogens.

PanGIA is a metagenomics analysis platform designed for pathogen detection and microbial community profiling, offering improved accuracy and usability for non-expert users.

Vorpal constructs interpretable linear models using clustered degenerate K-mers to predict binary phenotypes in RNA viruses, demonstrating its utility in three virus clades.