Fig. 8
From: Deep learning for DNase I hypersensitive sites identification
The figure above shows the ratio of the training data to the non-extended data. Firstly, the DHSs is sorted according to the length of the fragments. The vertical axis indicates the length of each fragment. The horizontal axis indicates the number of the fragment in the entire dataset. The blue line indicates the non-extended data, and the area underneath it indicates the number of bases in the whole non-extended data. The red dotted line indicates the training data, whose bottom area indicates the base number of the whole training data. The gray area indicates the extra fragments, which is also used to speed up training and extended by us