Skip to main content
Fig. 1 | BMC Genomics

Fig. 1

From: DeepHistone: a deep learning approach to predicting histone modifications

Fig. 1

Diagram of DeepHistone. The deep neural network model consists of three modules: a DNA module, a DNase module, and a joint module. The DNA module extracts sequence information via a densely connected convolutional neural network. The DNase module deals with chromatin accessibility information using the similar architecture. The joint module combines both sequence and chromatin accessibility features to distinguish histone modification sites of a marker from those of other markers. In figure, (128,4,9) indicates there are 128 kernels, each of width 4 and length 9 in a convolutional layer and (1,4) indicates kernel width and length are set to 1 and 4 respectively in a pooling layer. Others have similar meaning

Back to article page