D3.2 - Integrative analysis of multiple data types within cancer portals

This document describes the development of functionality for integrated analysis of cancer data for different standardised data types, including clinical data, radiomic images, time series data and DNA variant allelic frequency data, all in a portable format for use in federated analysis settings across different sites hosting relevant cancer data. The portability of the integration is crucial to enable federation. The federation aspect caters to the handling requirements of the human sensitive data to comply with FAIR data practices. The implementation utilises well established cancer analytical solutions, and provides a bootstrap for further development and local adaptation. Examples are provided, along with tutorials and guides, to facilitate dissemination. One main feature that has been developed is that of linking the existing software called cBioPortal, which is used for analysis and visualisation of molecular data, to trigger analyses in predefined workflows outside of cBioPortal, in this case to the workflow engine called Galaxy, with automatic delivery of results back to cBioPortal of the analysis performed. This is realised using time series data on cancer mutations in a given tumor, based on sequencing at multiple time-points, followed by analysis of the existence of clones of cell in the tumor, through an open source third-party software, PyClone VI, and the provision of results back to cBioPortal for visualization in an integrated presentation.