We evaluated DIABLO with
- a simulation study showing that DIABLO is able to select correlated and discriminatory features
- four multi-omics studies, integrating four omics datasets and benchmarked DIABLO against existing component-based supervised and unsupervised methods. We showed that the multi-omics biomarker signatures from DIABLO led to superior biological enrichment .
- two multi omics case studies in breast cancer and asthma, highlighting the versatility of DIABLO for various experimental designs (single point, repeated measurements, pathway module based).
All our R codes and analyses are available at the following link. The R code will be rendered on our website in the coming weeks.