Vladimir Gorshkov

DeepLC uses transfer learning to improve LC retention time prediction across different experimental setups and peptide modifications.

This paper introduces an ultrafast metaproteomics method using Orbitrap MS to rapidly and accurately assess microbial identity and functional activity without peptide fragmentation.

This paper evaluates the efficiency of negative ion mode proteomics using an MS1-only approach and finds it can identify over 1,000 proteins in human cells with high accuracy.

This study compares DirectMS1Quant, an MS1-only quantitation method, with MS/MS-based methods like LFQ, TMT, and DIA, showing comparable efficiency despite shorter runtime.

FLASHIda improves proteoform identification in top-down proteomics by intelligently selecting precursors, doubling identification counts in E. coli lysates.

The paper improves the DirectMS1 proteomics method using machine learning to identify over 2000 proteins in 5 minutes with 1% FDR.