Table 1. Comparison of predictivity of various modalities for each dataset/task. Predictivity is measured by the area under ROC curve (AUC). AUC values listed in the table were optimized over uni-modal methods given in Table S3. The highlighting of each cell corresponds to relative predictivity for each dataset/task – the more predictive is modality for a given dataset/task, the darker is its highlighting.

	TCGA_BRCA1.R1	TCGA_BRCA1.R2	TCGA_BRCA1.R3	TCGA_BRCA1.R4	TCGA_BRCA1.R5	TCGA_BRCA1.R6	TCGA_BRCA1.R7	TCGA_BRCA1.R8	TCGA_BRCA2.R1	TCGA_BRCA2.R2	TCGA_BRCA2.R3	TCGA_BRCA2.R4	TCGA_BRCA2.R5	TCGA_BRCA2.R6	TCGA_BRCA2.R7	TCGA_BRCA2.R8	TCGA_OVCA.R1	TCGA_OVCA.R2	TCGA_OVCA.R3	TCGA_OVCA.R4	TCGA_OVCA.R5	TCGA_OVCA.R6	TCGA_OVCA.R7	TCGA_OVCA.R8	TCGA_OVCA.R9	TCGA_OVCA.R10	TCGA_OVCA.R11
Clinical	0.590	0.656	0.646	0.646	0.633	0.630	0.725	0.594	0.578	0.587	0.697	0.689	0.604	0.806	0.759	0.721	0.667	0.690	0.657	0.543	0.650	0.577	0.582	0.549	0.614	0.639	0.652
Gene Expression	0.661	0.637	0.612	0.979	0.969	0.836	0.606	0.764	0.661	0.672	0.721	0.935	0.910	0.860	0.860	0.750	0.744	0.768	0.811	0.734	0.670	0.618	0.608	0.566	0.609	0.714	0.763
Protein Expression	0.650	0.743	0.733	0.983	0.953	0.812	0.713	0.733	0.584	0.618	0.705	0.936	0.881	0.845	0.644	0.716	0.695	0.687	0.688	0.704	0.665	0.592	0.610	0.581	0.602	0.738	0.757
Somatic Mutations	0.589	0.618	0.668	0.845	0.801	0.748	0.636	0.691	0.651	0.659	0.627	0.709	0.716	0.796	0.744	0.759	-	-	-	-	-	-	-	-	-	-	-
DNA Methylation	0.695	0.637	0.619	0.983	0.935	0.753	0.681	0.697	0.680	0.688	0.723	0.921	0.821	0.856	0.883	0.713	0.728	0.772	0.745	0.707	0.670	0.642	0.575	0.627	0.590	0.675	0.682
miRNA Expression	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	0.761	0.723	0.741	0.803	0.698	0.611	0.584	0.593	0.601	0.658	0.742
Copy Number	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
Tumor Imaging	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
GWAS	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-