Figure 1From: Active enhancer positions can be accurately predicted from chromatin marks and collective sequence motif dataComparison of prediction quality from histone marks. Difference in prediction quality achieved with BNFinder on epigenetic features for dataset of different sizes: 64 examples from [16] - AUC of0.75 on average (a) and 8008 examples from [17] - AUC of 0.93 on average (b). Both experiments are reported for cross-validated training.Back to article page