Abstract
Immunoinformatics approach has been used to identify potential T cell epitopes from structural and non-structural proteins for immunotherapy against novel coronavirus 2019-nCoV across populations Two different prediction algorithms, NetCTLpan and Pickpocket were used to generate consensus epitopes against HLA supertypes. All of the 57 epitopes identified had no similarity/identity with the human proteome thus preventing crossreactivity. Many of these epitopes formed a tight cluster around consensus sequences
MGYINVFAFPFTIYSLLLC
and KVSIWNLDYIINLI across proteins and alleles. These should be urgently tested in in-vitro MHC binding and T cell assays before being tried as vaccines to further prevent pandemic due to this lethal coronavirus.