Artificial intelligence for proteomics and biomarker discovery

Abstract

There is an avalanche of biomedical data generation and a parallel expansion in computational capabilities to analyze and make sense of these data. Starting with genome sequencing and widely employed deep sequencing technologies, these trends have now taken hold in all omics disciplines and increasingly call for multi-omics integration as well as data interpretation by artificial intelligence technologies. Here, we focus on mass spectrometry (MS)-based proteomics and describe how machine learning and, in particular, deep learning now predicts experimental peptide measurements from amino acid sequences alone. This will dramatically improve the quality and reliability of analytical workflows because experimental results should agree with predictions in a multi-dimensional data landscape. Machine learning has also become central to biomarker discovery from proteomics data, which now starts to outperform existing best-in-class assays. Finally, we discuss model transparency and explainability and data privacy that are required to deploy MS-based biomarkers in clinical settings.

Progress of proteomics‐driven precision medicine: From a glycosylation view

“…74 Recently, Cao et al 75 further development. 77,78 Although currently glycoproteomics was mainly applied to cancer analysis, there is also a bright future for glycosylation exploration in other diseases, such as neurodegenerative diseases and metabolic diseases.…”

“…In response to these demands and challenges, further improvements in technological approaches are required, especially in glycopeptides enrichment methods, effective ways of MS detections, strategies for interpreting site‐specific glycoforms, effective sample preparation toward various sample types, and methods for single‐cell analysis. Besides, as precision medicine is bound to rely on large data sets for accurate diagnosis and treatment, artificial intelligence techniques including traditional machine learning and deep learning are of great importance to assist further development 77,78 . Although currently glycoproteomics was mainly applied to cancer analysis, there is also a bright future for glycosylation exploration in other diseases, such as neurodegenerative diseases and metabolic diseases.…”

Proteomics in antibiotic resistance and tolerance research: Mapping the resistome and the tolerome of bacterial pathogens

“…In fact, machine learning has now become increasingly important for biomarker discovery from proteomics data [167,168]. This means that hunting for new clinically important antibiotic tolerance/resistance targets and markers would be easier, more accurate, and reliable.…”

“…For instance, deep learning can now be called upon to predict experimental peptide measurements from amino acid sequences alone, which could improve the quality and reliability of analytical workflows [167]. In fact, machine learning has now become increasingly important for biomarker discovery from proteomics data [167, 168]. This means that hunting for new clinically important antibiotic tolerance/resistance targets and markers would be easier, more accurate, and reliable.…”

Metabolic dyshomeostasis induced by SARS-CoV-2 structural proteins reveals immunological insights into viral olfactory interactions

“…It has been previously demonstrated that sophisticated machine learning (ML) approaches are useful workflows to integrate multi-omics data with the aim to discover complex interconnections between different type of entities (Ebrahim et al, 2016;Mann et al, 2021). To uncover regulatory complex composed by molecular effectors potentially modulated by SARS-CoV-2 structural proteins at olfactory level, transcriptomic and proteomic datasets were integrated and mined by a machine-learning approach in which multiple entities (genes, proteins, upstream regulators, pathways, diseases) were interconnected in the form of knowledge graphs.…”