Science
ALGORITHMS
The PIRCHE algorithms are specifically designed to estimate the indirect pathway of allo recognition. In brief, it comprises the following steps:
Antigen recognition
Internalization of mismatched allo HLA
Intracellular cleavage of HLA into peptides
Allo peptides are bound in HLA binding groove (pHLA)
Presentation of pHLA to extracellular space
pHLA specific T cell receptor activation induces T cell help
B cell proliferation and antibody production
Adverse clinical outcomes
scientific ROADMAP
T cell Epitope Matching
The high genetic variability of the HLA genome results in significant variance in peptide binding capability.
Artificial intelligence neural networks are utilized to identify HLA binding motifs, enabling the prediction of T cell epitope binding even for peptides that have not been experimentally tested.
These predictions are quantified by PIRCHE scores, representing the CD8 and CD4 T cell response.
Higher PIRCHE scores have been associated with adverse outcomes in diverse transplant scenarios.
B cell Epitope Matching
Although the donor HLA may differ from the patient HLA in only a few surface accessible amino acid positions, HLA antibodies are capable to distinguish and bind these antibody epitopes.
The Snow matching algorithm identifies surface-accessible amino acid mismatches in donor HLA to estimate antibody epitope incompatibility between patients and donors.
Snow utilizes two algorithms to analyze amino acid surface area and assess the three-dimensional structure of amino acids using a spherical mapping algorithm. The number of exposed mismatched amino acids is called the Snow score, relating to antibody epitope load.
