COVID-19 coronavirus vaccine design using reverse vaccinology and machine learning

Edison Ong; Mei U Wong; Anthony Huffman; Yongqun He

In this study, we first surveyed the existing coronavirus vaccine development status, and 98 then applied the Vaxign RV and Vaxign-ML approaches to predict COVID-19 protein 99 candidates for vaccine development. We identified six possible adhesins, including the structural 100 protein and non-structural proteins (including nsp3) using reverse vaccinology and machine 107 learning, we proposed and discussed a cocktail vaccine strategy, for rational COVID (Table 2) . 128 more typically mild HCoV-229E, HCoV-OC43, HCoV-NL63, and HCoV-HKU1. The sequence 138 conservation suggested the potential of N protein as a candidate for the cross-protective vaccine 139 against SARS and MERS. The N protein was also evaluated and used for vaccine development 140 (Table 2 ). The N protein packs the coronavirus RNA to form the helical nucleocapsid in virion 141 assembly. This protein is more conserved than the S protein and was reported to induce an 142 immune response and neutralize coronavirus infections 23 . However, a study also showed the 143 linkage between N protein and severe pneumonia or other serious liver failures related to the 144 pathogenesis of SARS 24 . 145 146 Six adhesive proteins in SARS-CoV-2 identified as potential vaccine targets 147