PEDLA: predicting enhancers with a deep learning-based algorithmic framework.

With 10,000 iterations of random shuffling of the enhancer predictions in the H1 genome, we found the average validation rate of distal DHS, p300 and TF to be 13.0%, 1.1% and 4.2%, respectively, and the actual validation rates were highly significant according to a one-sided t -test ( p -values = 0) ( Fig. 1B ).

also in 132,142 other papers