Spike protein binding prediction with neutralizing antibodies of SARS-CoV-2

Tamina Park; Sang-Yeop Lee; Seil Kim; Mi Jeong Kim; Hong Gi Kim; Sangmi Jun; Seung Il Kim; Bum Tae Kim; Edmond Changkyun Park; Daeui Park

The copyright holder for this preprint (which was not peer-reviewed) is the . https: //doi.org/10.1101 //doi.org/10. /2020 Table S1 ) was used as an input for ConSurf. In ConSurf, conservation scores and confidence 107 intervals for the conservation scores were calculated using the empirical Bayesian method. The scores were 108 normalized using the number of inputted sequences. Also, the highest score of ConSurf program means the most 109 conserved position among sequences. We additionally checked the epitope positions on the SARS-CoV-2 S 110 protein based on the known epitope information of 11 neutralizing antibodies developing for SARS-CoV and 111 MERS-CoV. Each information of epitope positions was acquired from literatures (Table 1) The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10. 1101 /2020 antibodies to prevent MERS-CoV were selected in the study (Table 1 ). The complex structure of RBD and ten 124 neutralizing antibodies was retrieved from PDB. The complex structures were superimposed to the RBD 125 structure of SARS-CoV-2 which were built by homology modeling. The procedures were performed that the 126 RBD structures of SARS-CoV2 and SARS-CoV were aligned by pairwise sequence alignment. And then the 127 structures were superimposed according to those pairwise alignments using MatchMaker program [29] . Finally, 128 we suceesfully predicted the complex structures of neutralizaing antibody candidates and RBD of SARS-CoV-2.