Artificial intelligence predicts the immunogenic landscape of SARS-CoV-2: toward universal blueprints for vaccine designs

Brandon Malone; Boris Simovski; Clement Moline; Jun Cheng; Marius Gheorghe; Hugues Fontenelle; Ioannis Vardaxis; Simen Tennoe; Jenny-Ann Malmberg; Richard Stratford; Trevor Clancy

In order to identify epitope hotspots that have the potential to be viable immunogenic targets for the vast majority of the human population, we first carried out a Monte Carlo random sampling procedure, on the epitope maps generated previously (for the Wuhan reference sequence exemplified in Figure 1 for the S-protein), to identify specific areas of the SARS-CoV-2 proteome that have the highest probability of being epitope hotspots (See Material and Methods). Three bin sizes were investigated for potential epitope hotspots; 27, 50 and 100. A statistic was calculated for each defined subset region of the protein (bin) from the set of 100 HLAs. The Monte Carlo simulation method was then used to estimate the p-values for each bin, whereby each bin represented a candidate epitope hotspot. The statistically significant bins that emerged from the simulation represented epitope hotspot or regions of interest for each protein analyzed. Epitope hotspots are built on the individual epitope scores, epitope lengths, and for each amino acid that they comprise. These scores are generated for each amino acid in the hotspots for all of the 100 HLA alleles most frequent in the human population. Based on the Monte Carlo analysis, the significant hotspots are those below a 5% false discovery rate (FDR), and represent regions that are most likely to contain viable T cell driven vaccine targets that can be recognized by multiple HLA types across the human population. A summary of the epitope hotspots identified across the entire spectrum of the virus is depicted in Figure 3 and reveals that the most immunogenic regions of the virus, that target the most frequent Human HLA alleles in the global population, are found in several of the viral proteins above and beyond the antibody exposed structural proteins, such as the S-protein.