with this initial gene cluster, and then rerun using the top hits (by s.g.i, the Gene Recommender normalized correlation metric) from the initial run. If the seed gene scores highly after the second run (i.e., within the top 50 genes most correlated with the putative cluster) LOOCV is used to trim the cluster to only the highest scoring hits (regardless of gene set size). Once a tightly regulated cluster has been found, the next-highest scoring genes are added incrementally while stringently keeping LOOCV scores low. Ultimately, all genes within a predicted cluster will have an approximately equivalent contribution to the cluster, as determined by LOOCV.
