Computational Prioritization of T Cell Epitopes to Overcome HLA Restriction and Antigenic Diversity in .

Developing a highly effective malaria vaccine remains challenging due to 's antigenic diversity and human leukocyte antigen (HLA) polymorphisms, which complicate vaccine antigen selection and limit immune protection. The first recommended malaria vaccine, RTS,S, provides only partial, allele-specific protection with waning immunity over time, and the more recently developed R21 vaccine will likely encounter the same hurdles. To address these challenges, we developed a computational tool that integrates sequence diversity, predicted T cell epitope-HLA binding affinities, and HLA allele frequencies from sub-Saharan Africa to identify conserved, immunogenic epitopes with broad population coverage. We analyzed 42 proteins, previously identified as vaccine candidate antigens, and generated consensus sequences using data from 18 African countries, and then incorporated HLA allele frequencies from 24 sub-Saharan populations. CD8+ and CD4+ T cell epitopes were predicted using NetMHCpan-4.1 and NetMHCIIpan-4.1. Our novel tool, T cell Epitope Nomination (TEpiNom), used greedy optimization to filter and select epitopes based on epitope sequence conservation (>95%), binding affinity (median rank <10%), and broad HLA coverage, minimizing redundancy to reduce immune escape risk. Our tool identified 2,265 MHC I and 1,992 MHC II conserved epitopes spanning pre-erythrocytic, erythrocytic, and sexual stage proteins. Key MHC I epitopes from pre-erythrocytic antigens HSP70-2, SLARP/SAP1, p36, FabZ, LISP1, LSA1, UIS3, p24_2, PL, and FabG achieved near 100% HLA-A, HLA-B, and HLA-C coverage, and MHC II epitopes from pre-erythrocytic, erythrocytic, or sexual antigens provided 98.5%-100% coverage for a given parasite life stage. This strategy advances malaria vaccine design by integrating epitope promiscuity and multistage antigen selection to support broad, durable protection and identify promising multi-epitope malaria vaccine candidates for subsequent experimental validation. Our computational framework is adaptable for vaccine development against other genetically diverse and immunologically evasive pathogens.