Skip to main content
. 2017 Aug 15;7:8145. doi: 10.1038/s41598-017-08414-y

Figure 1.

Figure 1

Generation of cognate and non-cognate antibody–antigen complexes. The starting structures for cognate models (blue) are the crystal complexes, while the starting non-cognate models are generated by superimposing the VL–VH framework of the non-cognate antibody (red) on the cognate antibody in the crystal complex. A customized version of RosettaDock’s high-resolution stage is then used to generate an ensemble of models focused around the epitope for discrimination.