Zhiwei Qin

The paper uses a protein language model to analyze the evolutionary history of bacterial aromatic polyketide ketosynthases and chain length factors, identifying a shared ancestral cluster and distinct evolutionary paths.

DeepAden predicts A-domain substrate specificities in NRPSs using a two-stage deep learning framework, achieving competitive performance and enabling the identification of novel NRP gene clusters.

This study identifies holomycin as a key factor in Yersinia ruckeri infection of rainbow trout at low temperatures, showing it remodels host microbiota and enhances biofilm formation.

MAAPE integrates KNN and co-occurrence matrix analysis to extract evolutionary insights from protein embeddings, successfully reconstructing phylogenetic relationships in benchmark datasets.

DNA-SaM improves DNA data storage by achieving high efficiency, scalability, and safety through advanced encoding and architecture.

This study uses machine learning-guided directed evolution to develop a peptide antibiotic from antimicrobial peptides, showing improved antimicrobial activity and therapeutic potential.

The paper reports new aromatic polyketides called formicapyridines, shows how mutating a single gene increases their production, and identifies a non-catalytic role for ABM superfamily proteins in maintaining biosynthetic pathway fidelity.