Related Experiment Video
Updated: Jun 19, 2026
A High Throughput MHC II Binding Assay for Quantitative Analysis of Peptide Epitopes
Published on: March 25, 2014
TCEPVDB: Artificial Intelligence-Based Proteome-Wide Screening of Antigens and Linear T-Cell Epitopes in the
Mansi Dutt1,2,3, Anuj Kumar1,2,3, Ali Toloue Ostadgavahi1,2
1Department of Microbiology and Immunology, Canadian Center for Vaccinology (CCfV), Faculty of Medicine, Dalhousie University, Halifax, NS B3K 6R8, Canada.
Background:
Poxviruses constitute a family of large dsDNA viruses that can infect a plethora of species including humans. Historically, poxviruses have caused a health burden in multiple outbreaks. The large genome of poxviruses favors reverse vaccinology approaches that can determine potential antigens and epitopes. Here, we propose the modeling of a user-friendly database containing the predicted antigens and epitopes of a large cohort of poxvirus proteomes using the existing PoxiPred method for reverse vaccinology of poxviruses.
Methods:
In the present study, we obtained the whole proteomes of as many as 37 distinct poxviruses. We utilized each proteome to predict both antigenic proteins and T-cell epitopes of poxviruses with the aid of an Artificial Intelligence method, namely the PoxiPred method.
Results:
In total, we predicted 3966 proteins as potential antigen targets. Of note, we considered that this protein may exist in a set of proteoforms. Subsets of these proteins constituted a comprehensive repository of 54,291 linear T-cell epitopes. We combined the outcome of the predictions in the format of a web tool that delivers a database of antigens and epitopes of poxviruses. We also developed a comprehensive repository dedicated to providing access to end-users to obtain AI-based screened antigens and T-cell epitopes of poxviruses in a user-friendly manner. These antigens and epitopes can be utilized to design experiments for the development of effective vaccines against a plethora of poxviruses.
Conclusions:
The TCEPVDB repository, already deployed to the web under an open-source coding philosophy, is free to use, does not require any login, does not store any information from its users.
