Enrichment of miR-9 seed sites in 3′ end untranslated regions (3′UTRs). (A) Enrichment profile for the top ten enriched words in 3′UTRs after miR-9 transfection. Each line represents the running sum over all scores that quantify degree of enrichment according to gene downregulation (from most downregulated to most upregulated). miR-9 seed sites (7mer-m8 seed ACCAAAG) and similar words dominate the top ten regulatory words predicted. (B) Word cluster plot showing words in 3′UTRs ranked by differential expression, after miR-9 transfection. Each dot represents a word, summarizing Z-scores, and enrichment specificity indices of the enrichment profiles of negatively correlated 6, 7 and 8mer words. Triangles annotate known seed sites of human miRNAs. Triangles with a black border show miR-9 seed sites. Words are clustered by sequence similarity using the UPGMA algorithm and colored according to what motif (or cluster) they belong to (only some among top 100 words are plotted).