PMC

Graphs from miRNA–miRNA correlation with the same criteria as in but for nonprogressed subjects. This graph is similar to that in . miRNA, microRNA.

Graphs from miRNA–miRNA correlations. Edges represent strong correlations. Looped regions are common subgraphs. Networks shown are strongly correlated miRNAs among unaffected controls. miRNA, microRNA.

Graphs from miRNA–miRNA correlations with the same criteria as in and but for progressed subjects. Evidently much organisation of miRNA networks is lost in subjects who eventually progressed to psychosis. miRNA, microRNA.

Histogram of one AUC from true data vs. 1000 AUCs of classifiers built by the same greedy algorithm applied to pseudo data (NP and P labels randomly permuted). Fitted with a beta distribution, the AUC from real data indicates a P-value of 0.012. Since 17 random AUCs of 1000 by chance exceed the true AUC, an algebraic method gives alternative P-value=0.018. Thus, the performance of the Greedy Algorithm limited to selection of at most six markers and applied to the full data set is unlikely to be chance. AUC, area under the curve of receiver operating characteristic.

(a) The Greedy Algorithm was applied to 1000 selections of random 80% subsets of nonprogressed subjects and random 80% subsets of progressed subjects. Each time up to six markers could be chosen. The seven most frequently chosen markers are shown with their selection rates. The solid bars indicate the five markers that were also selected in the first six markers chosen by the Greedy Algorithm for the full data set (), yielding as a sum of z-scores the classifier function in sum (). This function applied to the full data yields AUC=0.86. (b) ROC of the five-miRNA classifier function (). Dotted lines are 95% confidence levels, and the dashed line is hypothetical performance of a random classifier. AUC, area under the curve; miRNA, microRNA; ROC, receiving operating characteristics.