Fig. 5From: Integrative, multi-omics, analysis of blood samples improves model predictions: applications to cancerVariable importance plot from random forest on cancer subtype in the TCGA dataset. First ten variables ranked by variable importance (in terms of mean Gini index) in the full integrative model for cancer subtype in the TCGA dataset. \(\texttt {Joint}_i\) denotes the i-th joint component estimates by aJIVE, while \(\texttt {MetInd}_i\), \(\texttt {mRNAInd}_i\) and \(\texttt {miRNAInd}_i\) are the i-th individual components estimated by aJIVE for methylation, mRNA and miRNA respectivelyBack to article page