TFBS motif hits in pre-miRs as compared to permuted negative control sets. Human pre-miR sequences were scanned directly for predicted transcription factor binding sites using matrices from the TRANSFAC 7.0 Public Database at the indicated stringency (z-score ≥ 1.64 or 2.33, see Methods). For each pre-miR, a randomized negative control set was created consisting of 1,000 iterations of permuted sequences matched for length and dinucleotide frequency. Each of the randomized sequences was then scanned for TRANSFAC motif hits in the same manner. **significant at p < 3 × 10-5. ***p < 2 × 10-12.