Daniel Jimenez

MicrobeRT is a framework of fine-tuned language models for metagenomic analysis, achieving strong performance in taxonomic classification, antibiotic resistance detection, and pathogenicity assessment.

This paper compares generative models for designing non-oxime reactivators of paraoxon-inhibited acetylcholinesterase, finding that down-selection methods outperform DrugDiff in practical drug-likeness and synthesis feasibility.