Premium
Front Cover: DeepRescore: Leveraging Deep Learning to Improve Peptide Identification in Immunopeptidomics
Author(s) -
Li Kai,
Jain Antrix,
Malovannaya Anna,
Wen Bo,
Zhang Bing
Publication year - 2020
Publication title -
proteomics
Language(s) - English
Resource type - Reports
SCImago Journal Rank - 1.26
H-Index - 167
eISSN - 1615-9861
pISSN - 1615-9853
DOI - 10.1002/pmic.202070151
Subject(s) - deep learning , identification (biology) , cover (algebra) , artificial intelligence , front cover , computer science , machine learning , engineering , biology , ecology , mechanical engineering
DOI: 10.1002/pmic.201900334 Sensitive and reliable peptide identification from immunopeptidomics data remains a major challenge primarily due to the inflated search space caused by the absence of enzymatic digestion in immunopeptidomics experiments. In article number 1900334, Kai Li et al. present DeepRescore (illustrated as a magnet in the cover art), a post‐processing tool to address this challenge. In DeepRescore, a machine learning model with features derived from deep learning predictions and other features is used to rescue good peptide‐spectrum matches (illustrated as cans) from the ones discarded in the initial analysis (illustrated as garbage). The performance improvement by DeepRescore is, to a large extent, driven by features generated from deep learning models AutoRT and pDeep2 (illustrated as two AI robots). The cover was designed by Miaoyu Li.