Artificial Intelligence (AI) based approach for identifying biomarkers associated with neutralizing antidrug antibodies to Factor VIII used in Hemophilia A treatment

Abstract Lack of clinically validated markers associated with the development neutralizing antibodies (inhibitors) to replacement protein-drug, Factor VIII (FVIII) is a major challenge in managing Hemophilia A (HA). By utilizing Machine Learning (ML) and Explainable AI (XAI) we identified ranked relevant variables (biomarkers) that could be predictive for developing inhibitors FVIII drugs hemophilia patients. The dataset used study was derived from My Life Our Future (MLOF) repository included such as age, race, ethnicity, mutations F8 gene Human Leucocyte Antigen Class II (HLA-II) type. We computed number foreign peptides, based on alignment endogenous infused drug sequences, foreign-peptide HLA-II molecule binding affinity (using NetMHCIIpan algorithm). complete added trained using multiple ML classification models identify top performing model, which chosen further processing XAI via SHAP (SHapley Additive exPlanations) critical prediction inhibitor five predicting values are: (i) baseline activity protein, (ii) mean all peptides HLA DRB 3, 4, & 5 alleles, (iii) DRB1 (iv) maximum among (v) mutation This research supported by funds Hemostasis Branch/Division Plasma Protein 344 Therapeutics/Office Tissues Advanced Therapies/Center Biologics Evaluation Research 345 U.S. Food Drug Administration part an appointment Participation 346 Program at Center administered Oak Ridge Institute 347 Science Education through interagency agreement between Department Energy 348 Administration. MLOF program developed partnership 349 NHF, ATHN, Bloodworks Northwest, Bioverativ financially 350 Bioverativ, ATHN.