Deep learning neural network tools for proteomics

• Published in: Cell reports methods Vol. 1; no. 2; p. 100003
• Main Author: Meyer, Jesse G.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 21.06.2021; Elsevier
• Subjects: bioinformatics; deep learning; mass spectrometry; MS/MS; neural networks; peptides; proteomics; retention time; Review; deep learning; retention time; proteomics; neural networks; bioinformatics; mass spectrometry; peptides; MS/MS
• ISSN: 2667-2375; 2667-2375
• DOI: 10.1016/j.crmeth.2021.100003

Mass-spectrometry-based proteomics enables quantitative analysis of thousands of human proteins. However, experimental and computational challenges restrict progress in the field. This review summarizes the recent flurry of machine-learning strategies using artificial deep neural networks (or “deep learning”) that have started to break barriers and accelerate progress in the field of shotgun proteomics. Deep learning now accurately predicts physicochemical properties of peptides from their sequence, including tandem mass spectra and retention time. Furthermore, deep learning methods exist for nearly every aspect of the modern proteomics workflow, enabling improved feature selection, peptide identification, and protein inference. In this review, Meyer summarizes and contrasts the different machine-learning strategies that use neural networks for (1) prediction of peptide properties from sequence and (2) peptide/protein identification. Limitations and opportunities of these deep learning tools in mass-spectrometry-based proteomics are also discussed.