Skip to main content
. Author manuscript; available in PMC: 2017 Jul 7.
Published in final edited form as: Res Comput Mol Biol. 2017 Apr 12;10229:336–352. doi: 10.1007/978-3-319-56970-3_21

Table 3.

Comparison between 6-mer OLS models and PWM models for four TFs with MPRA data available from [15]. The numbers represent areas under the receiver-operating characteristic (ROC) curve, for false positive rates between 0 and 0.2. The maximum value for the area under the curve is 0.2. The expected value for random models is 0.02. Results show that 6-mer OLS models outperform PWM models for 3 out of the 4 TFs tested.

Hnf1 Hnf4 Gata Foxa
PWM models 0.04556 0.02082 0.07014 0.04835
6-mer OLS models 0.06667 0.04868 0.05046 0.05275