Figure 1

Error rates of distinguishing type I versus type II TF binding sites in synthetic ChIP-seq data sets by using a serial version of BayesPI2+. Error rates (i.e. TP - true positive rate, FP - false positive rate, TN - true negative rate, and FN - false negative rate) are displayed in the box plot, which were estimated by randomly splitting training and testing data multiple times. The inferred best TF PBEM from training data was used to compute dbA and to distinguish type I versus type II TF binding sites in the testing data.