Pan-cancer classification of multi-omics data based on machine learning models

AbstractThe integration of multiple biological layers derived from different omics studies generates a novel concept of pan-cancer molecular classification, suggesting new therapeutic strategies for precision medicine. In this review, we will present a comprehensive portrait of the latest advances for multi-omics combination in oncology considering different cancer types. We will show the different applications of machine learning for characterizing cancer biology and the identification of prognostic and response to therapy prediction opening the scenario to personalized therapy. We grouped the selected articles into six main applications: (1) response to therapy, (2) survival prediction, (3) cancer driver genes prediction, (4) predictive biomarker (5) biological process, and (6) cancer subtype prediction. This review suggests that the integration of multi-omics data can improve cancer knowledge showing a good performance of the results with machine learning methods.
Source: Network Modeling Analysis in Health Informatics and Bioinformatics - Category: Bioinformatics Source Type: research