Predictive, preventive, and personalized medicine is the future of cancer research. Cancer is a complex, whole-body disease that involves multiple factors, multiple processes, and multiple consequences. A series of molecular alterations at different levels of genes (genome), RNA (transcriptome), proteins (proteome), peptides (peptidome), metabolites (metabolome), and imaging characteristics (radiomics) that result from exogenous and endogenous carcinogens are involved in tumorigenesis and mutually associate in a network system, thus making it difficult to use a single molecule as a biomarker for personalized prediction, prevention, diagnosis, and treatment of cancer.

Previous studies have identified multiple potential prognostic signatures with remarkable clinical efficacy in cancer management based on omics data such as transcriptomics, proteomics, and epigenomics. Comprehensive research that integrates biomarkers, pathological features, and imaging signatures will be significant. It is expected that the development of systematic prognostic models will enhance disease diagnostics and promote clinical management. Machine learning has been applied to multiple research areas due to its ability to process large-scale data, identify common features of different classifications, and offer guidance for clinical decisions. Therefore, building prognostic models based on machine learning approaches has received increasing attention. Recently in medicine, machine learning has been shown to assist with alternative splicing prediction, drug sensitivity scrutiny, patient survival outcome prediction, tumor diagnosis, and tumor classification. Adopting machine learning to identify prognostic factors in omics data – such as genomics, transcriptomics, proteomics, peptidomics, metabolomics, and radiomics – and calculate corresponding prognostic models gives a more precise prognostic prediction.

This Special Issue will focus on prognostic models based on machine learning for cancer research. We welcome original research as well as review articles.

Potential topics include but are not limited to the following:

• Prognostic models based on clinical features such as histopathological slides, radiomics, or other factors
• Prognostic models involved in different types of tumors, pan-cancer analysis, or data from a variety database
• Machine learning-related prognostic prediction and drug sensitivity prediction
• Prognostic models based on traditional omics data, including mRNA, non-coding RNA, DNA methylation, etc.
• Verifying results on clinical samples or cell lines, or adopting multiple cohorts including training and verification cohorts