Constructing global HLA immunodominance hierarchies


Despite the large number of epitopes that a HLA allele can potentially bind to, in many cases, a single HLA-epitope pair is sufficient to elicit an effective CTL response and such epitopes are considered to be immunodominant. Such immunodominant HLA-epitope pairs, such as B*07:02-NP, that are known to offer higher protection against severe disease (Peng et al., 2021), have been well documented in case of COVID-19. We discuss the possibility of similar immunodominance patterns among HLA alleles in a population that can be extrapolated to analyse such patterns at the individual level. This would allow us to prioritize epitopes recognized by immunodominant alleles in an individual and also identify protective and susceptible alleles at global and population levels.

Proposed workflow

In order to arrive at this hierarchy, we plan to construct a network with nodes representing HLA alleles and edges representing the nature of immunodominance between the two nodes (alleles). To characterize immunodominance, we compare the frequency of association of a HLA allele with a particular disease severity in presence and absence of another HLA allele. If a significant difference is observed between the compared frequencies, an edge is drawn, directed towards the allele whose presence altered the frequency of association of the former allele. If the introduction of the new allele increased protection, we assign a positive edge weight and if it decreased protection, we assign a negative edge weight. This method of edge assignment could result in unidirectional as well as bidirectional edges (where frequencies of association of a pair of HLA alleles with severity states significantly differs from their individual frequencies). Once we have a large sample set covering individuals with diverse set of HLA alleles in a population, the network can be expanded to cover immunodominance patterns within a cohort. Availability of HLA data from diverse populations would then enable construction of a global HLA immunodominance network.

Potential applications

  • Identification of immunodominant HLA alleles: Nodes with only incoming edges without outgoing edges would represent the most immunodominant alleles at a cohort or global level.

  • The network can also be used as a reference for personalized identification of immunodominant HLA alleles. Once the HLA genotype of an individual is obtained, the alleles can be mapped onto the immunodominance network and a ranked list of HLA alleles based on their immunodominance can be created. Identification of such immunodominant HLA alleles within an individual would be followed by identification of strong binding CTL epitopes recognized by these alleles, using bioinformatic epitope prediction tools. This would be followed by experimental validation of a strong CTL response elicited by the identified immunodominant HLA-epitope pairs.

  • Another potential application involves identification of high-risk individuals based on the immunodominance ranks corresponding to the expressed HLA alleles.

Data requirement

In order to implement this, we would require HLA genotype data of COVID-19 patients along with the severity of disease they faced. Ideally, we aim to maximize coverage of HLA alleles found throughout the world. To achieve this, we plan to construct smaller immunodominance hierarchies within cohorts followed by integration into a global hierarchy.

Overall, our project provides a strong foundation for construction of a global HLA immunodominance hierarchy that can be potentially utilized to identify immunodominant HLA alleles and epitopes, providing further support to T-cell based vaccine strategies against SARS-CoV-2. We would like to invite researchers who would be interested in sharing the relevant data detailed above, and collaborating on this project. Please feel free to contact us via email, if interested (contact details provided below). 

Contact information:

Vishal Rao
Undergraduate research student
Chandra Lab, Department of Biochemistry
Indian Institute of Science
Bangalore-560012, India

P.I Contact information:

Nagasuma Chandra, PhD
Department of Biochemistry
Indian Institute of Science
Bangalore-560012, India


Peng, Y., Felce, S. L., Dong, D., Penkava, F., Mentzer, A. J., Yao, X., … & Dong, T. (2022). An immunodominant NP105–113-B* 07: 02 cytotoxic T cell response controls viral replication and is associated with less severe COVID-19 disease. Nature immunology, 23(1), 50-61.