*Purpose: Functional epitopes, or eplets, represent a set of amino acids within three angstroms of one another that may not be sequentially contiguous but are neighbors in three-dimensional space. The HLA Matchmaker tool, developed by Rene Duquesnoy, catalogues the eplets of each HLA allele and identifies mismatched eplets between donor:recipient pairs. The degree of eplet mismatching correlates with patient outcomes. Unfortunately, the scalability and magnitude of application has been restricted by the time and effort required at the user interface

*Methods: We developed a software package using the programming language R, with capacity to build into a user-friendly web application, that dramatically reduces the effort of the end user. The tool is a compilation of several functions. HLA typing data on a population of one or many donors and recipients is read into R. The data is then converted into a consistent format using the function AlleleClean. Cleaned data are processed using the functions, CalEpletMHCI and CalEpletMHCII, which generate both a detailed output of the eplet mismatches between paired samples and a simplified output with the numerical count of mismatched eplets. Allele level mismatch can be analyzed with the function EvalAlleleMism. Finally, the AlleleTopN function generates a list of the most common alleles in the user’s dataset. We demonstrate through applied vignettes the simplified workstream allowed by our R package, hlaR, accessible at https://github.com/LarsenLab/hlaR.

*Results: User data input can be tailored to the intended application. The most common use of eplet data has been retrospective analysis of patient outcomes in the context of the mismatch load between recipient and donor (1 to 1). Analyzing a dataset of contrived donor recipient pairs with high-resolution HLA typed subjects, we observed equivalent results between hlaR and eplet mismatches calculated by HLAMatchmaker. Results were consistent excepting minor typographical errors traced to the excel workbook. To emulate the use of mismatch load in organ allocation, we provide an additional vignette, calculating mismatch between a single donor and a set of possible recipients or, similarly, between a set of donors and a single recipient (1 to many, many to 1).

*Conclusions: Our new tool, hlaR, can provide simplified eplet data with a streamlined workflow for multiple applications. We plan to expand the functionality of this package to include imputation of low-resolution data by incorporating a query of the National Marrow Donor Program’s HaploStats web application. With decreased effort from the end user, eplet matching and mismatch load data, which have been significantly associated with graft outcomes, can be further incorporated into both research and clinical use.

« Back to 2021 American Transplant Congress