David B. Li

Minerva extracts RNA base-pairing, protein-contact and repeat maps from a genome language model without alignments, flagging unannotated structured loci and a phage ncRNA array reverse-transcribed into short cDNAs.

This study determines the structural basis of phage terminase recognition by Avs2 and Upx, revealing distinct binding mechanisms despite structural differences.

This study demonstrates that Staphylococcus epidermidis can be engineered to elicit potent immune responses when applied topically or intranasally, with nasal colonization inducing stronger systemic and mucosal immunity.

Fine-tuned Evo 1 and Evo 2 genome language models designed complete ΦX174-like phage genomes; 16 of 285 assembled designs were viable, and some outperformed ΦX174 in competition.

A 40B-parameter DNA model trained on 9.3 trillion tokens scores coding and noncoding variants zero-shot, exposes interpretable sequence features, and writes chromosome-scale sequences.

7B-parameter convolution-attention DNA model with 131 kb context, trained on prokaryotic genomes, performs zero-shot fitness and essentiality prediction and generates Cas loci, transposons, and 650 kb sequences.

LAMP-Seq combines three-step isothermal RT-LAMP amplification with combinatorial barcoding to pool and sequence SARS-CoV-2 clinical samples en masse for scalable diagnostics.