Proteomics holds immense significance in fundamental and applied research in various fields, including life sciences, medicinal sciences, and pharmaceutical sciences. The rapid development of mass spectrometry (MS) technologies has facilitated MS-based proteomics research, which has emerged as one of the primary methods for determining the composition, structures, and functions of proteins. The necessity of processing these complex datasets has increased significantly owing to the growing volume and diversity of MS data pertaining to proteins. Artificial intelligence (AI) possesses powerful data processing abilities, and is being increasingly employed for handling these challenges. In particular, deep learning has been extensively employed in MSbased proteomics research. This review discusses and compares the different AI algorithms developed for various tasks, including the prediction of protein spectra, retention times, peptide sequences, and MS-based protein structure prediction, and highlights their respective strengths and weaknesses. The limitations and future prospects of AI in MS-based proteomics research are additionally discussed herein.
Keywords: Deep learning; algorithms; datasets.; peptide sequence; protein spectra; protein structure prediction; retention time.
Copyright© Bentham Science Publishers; For any queries, please email at epub@benthamscience.net.