DEVELOPMENT OF A MODULAR CONCEPT OF HLA TO ACHIEVE COMPREHENSIVE PEPTIDE BINDING PREDICTION.

6.5
#45-OR
DEVELOPMENT OF A MODULAR CONCEPT OF HLA TO ACHIEVE COMPREHENSIVE PEPTIDE BINDING PREDICTION.
David S. DeLuca MSc ¹, Barbara Khattab PhD ¹ and Rainer Blasczyk MD ¹. ¹ Department of Transfusion Medicine, Hanover Medical School, Hanover, Germany .

A variety of algorithms have been successful in predicting HLA-peptide binding for HLA variants for which plentiful experimental binding data exist. While predicting binding for only the most common HLA variants may provide sufficient population coverage for vaccine design, successful prediction for as many HLA variants as possible is necessary to understand the immune response in transplantation. However, the high cost of peptide sequencing limits the acquisition of binding data. Therefore, a prediction algorithm, which applies the binding information from well-studied HLA variants to HLA variants, for which no peptide data exist, is necessary. To this end, a modular concept of HLA-peptide binding prediction was developed. To verify the efficacy of this modular approach, accurate predictions were made for the following alleles without using experimental peptide binding data specific to those alleles: A*0201, A*0206, A*0214, B*2705, B*3501, B*5102, B*5301. Using the MHCBN peptide database, and a minimum cutoff of 15 peptides, the modular concept increased the number of predictable alleles from 15 (4.5%) to 75 (22.3%) of HLA-A and 12 (2.0%) to 36 (5.9%) of HLA-B proteins. Under the modular concept, binding data of certain HLA molecules can make prediction possible for numerous additional HLA alleles. In this regard, the HLA molecules A*7401, 3201, 6813; B*1803, 4103, 3908 have been identified to be the most informative. Achieving peptide binding prediction for all HLA molecules will provide a major basis for individualizing the clinical application of the minor histocompatibility concept and the management of Graft versus Leukemia reactions.