Skip to main content
. 2021 Feb 1;22:38. doi: 10.1186/s12859-020-03952-1

Fig. 6.

Fig. 6

Distribution of average normalized DNase coverage values of different regions with the inputs of JUND. The predictions from both models with and without attention from our training are evaluated by the true positive labels. Then the average normalized DNase coverage is calculated based on bins classified differently by the two models