Skip to main content
. 2020 Nov 28;37(10):1339–1344. doi: 10.1093/bioinformatics/btaa970

Fig. 2.

Fig. 2.

The performance of different DeepHiC models. (A) DeepHiC models were trained with sequences from bin pairs with length of 10, 40 and 100 kb, respectively. (B) The performance of the best DeepHiC model trained on bin pairs with 40 kb length was evaluated on independent datasets generated from GM12878 cells and K562 cells