Jason McDermott

VaLPAS pairs a bi-directional autoencoder with correlation for omics guilt-by-association with calibrated confidence; in L. starkeyi its autoencoder linked 176 unidentified metabolites to known-function proteins at ≥50% confidence.

This study identifies a novel E3 ligase secreted by Pseudomonas aeruginosa that regulates virulence through ubiquitin modification.

PathwaySeeker uses AI to reason about organism-specific metabolic networks using proteomic and metabolomic data, distinguishing confirmed reactions from hypotheses.

spammR is an R package for analyzing and integrating spatial multi-omic data, focusing on mass-spectrometry datasets and enabling cross-type integration within tissue samples.

The paper introduces an extension to the Snekmer tool that uses kmer libraries to annotate protein sequences with functional annotations, showing improved sensitivity over HMMs in metagenomic datasets.

The paper compiles a global soil virus dataset to assess biodiversity and biogeochemical impacts, finding most viral diversity remains unexplored and viruses may influence carbon and nutrient cycling.

MetaPathPredict predicts KEGG module presence in incomplete bacterial genomes using machine learning models.

The paper predicts ex vivo drug sensitivity in AML patients using proteomic and phosphoproteomic signatures, showing better accuracy than mutation or mRNA-based signatures.

The paper identifies 24,000 soil viral sequences and their hosts, revealing new viral diversity and roles in carbon metabolism across soil ecosystems.