Skip to main content
. 2017 May 24;10(Suppl 1):34. doi: 10.1186/s12920-017-0264-3

Fig. 3.

Fig. 3

Tissue-specific classifiers trained on 4-mers. a, b: Distribution of prediction values returned by heart- (a) and brain- (b) specific classifiers for different sets of sequences. c: Distribution of whole-genome prediction values (probability of being enhancer) for DNase-hypersensitive sites (DHS), and regions non-sensitive to DNase (non-DHS)