Figure 2.

The average KL divergence of feature distributions between (1) TF and PPI, and (2) TF and NonPF datasets for the GM and BM models, when the top ranked features are considered (TF& PPI_GM = feature distribution in TF vs. feature distribution in PPI in GM model, etc.)

Yang et al. BMC Bioinformatics 2008 9(Suppl 3):S11   doi:10.1186/1471-2105-9-S3-S11