Overview of the single-task pipeline.

<p>For genotype data, we associate each gene with a single SNP (a). Next, we calculate correlation statistics using the gene-based data for each data type (b). We then calculate enrichment scores using the correlation statistics for each data type (c). Finally, we independently build a predictive model for each data type using the enrichment scores for each data type and a standard SVM (d). In this overview, ASSESS corresponds to steps b and c.