Skip to main content
. Author manuscript; available in PMC: 2020 Mar 1.
Published in final edited form as: IEEE/ACM Trans Comput Biol Bioinform. 2017 Dec 18;16(2):596–606. doi: 10.1109/TCBB.2017.2784434

Fig. 6.

Fig. 6

A training example of the DCNN in ConLooper. a) The distance matrix of a template found by a 3rd party alignment tool. b) Input distance matrix for DCNN. It is a combination of 4 corner squares marked as black boxes in a). The boundaries of these black boxes are derived from the gap position in the structure. c) Output distance matrix for CNN. It is created from the white box in a).