Yuqi Zhang

DIA-BERT improves DIA proteomics analysis by using pre-trained transformer models, achieving higher protein and peptide identifications than DIA-NN in cancer samples.

DDA-BERT improves PSM rescoring in proteomics using a transformer-based model trained on 95 million human spectra, outperforming existing methods across species and sample types.

This study reveals structural and functional features of mammalian PIWI-piRNA-target complexes that enable efficient and broad target silencing.

The paper reports cryo-EM structures of DHHC9-GCP16 and Erf2-Erf4 complexes, revealing that GCP16 stabilizes DHHC9 and that phospholipid binding and specific palmitoylation are essential for catalytic activity.