"In silico Prediction of HLA-associated Drug Hypersensitivity"
Adverse drug reactions are a leading cause of morbidity and mortality with estimated annual inpatient costs of over $100 billion in the US alone. A growing number of the most severe adverse reactions, termed idiosyncratic drug hypersensitivities, are observed to be immune system mediated with genetic associations to specific human leukocyte antigen (HLA) alleles. However, little is known about the underlying mechanisms of the majority of these associations, which critically hinders preventative action in clinical settings. Previous work using in silico and in vitro approaches has demonstrated that the antiviral drug abacavir can bind the antigen binding cleft of HLA-B*57:01 and alter its specificity for self-peptides presented to T cells. A critical barrier to examining the generality of this model, termed "altered peptide repertoire model", is our inability to experimentally test all possible combinations between major HLA variants and multiple drugs. Here we describe computational approaches that can better focus such assessments. I will briefly describe previous work which combines homology modelling, molecular docking and molecular dynamics (MD) simulations to predict HLA-drug interactions. I will then describe current approach utilizing free energy perturbation based binding affinity calculations to improve the HLA-drug interaction rankings. The predictive power of our approach is tested on a set of drugs with known HLA-linked hypersensitivity reactions: abacavir with B*57:01, the gout prophylactic allopurinol with B*58:01, and the antiepileptic carbamazepine with B*15:02. Our studies represent a first step toward the development of a preclinical screening process that aims to identify drugs with a high risk of causing drug hypersensitivity.