Biological systems are complex entities involving complex interactions among different molecular compartments. Machine learning approaches can be used to build models that can help us to disentangle complex multi-omics relationships that are at the basis of complex diseases like cancer and find novel therapeutical targets.

The application of systems biology models to discover cancer related mechanism of action is powerful tool that has been successfully employed in the discover of key elements in the molecular setup of cancer and the development of resistance.

These methods are mainly based on the information derived from single omic assays while the integration of different omics layers is still an on field.

Multi omics models can provide a broader picture of the complex interactions behind cancer and can overcome many limitations of single omics-based models.

Given the increasing availability of public repositories containing multi omics assays of different cancer types for several patients annotated with clinical information (e.g. TCGA dataset), we ask if it is possible to build predictive molt omics based models to predict clinical outcomes of interest along with related multi-omics molecular mechanisms.