A computational pipeline for drug discovery in immunometabolism against autoimmune diseases

Abstract We developed a computational pipeline to study immunometabolism for drug discovery in immune mediated diseases. Our pipeline integrates RNA-seq and proteomics data to build genome scale metabolic models and integrate these models with external drug databases and patient data to perform metabolic flux analyses for identifying drug targets and repurposing. We applied our pipeline to explore the metabolism of B cells in the context of rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). We constructed metabolic models of B cells, performed enzymatic inhibition of targets of existing drugs and compounds, and integrated the model's results with differentially expressed genes in RA and SLE patients. Using a computational scoring method, we ranked the inhibited targets based on their ability to stimulate the reactions of differentially expressed genes in diseases to have metabolic flux in opposite directions of their up and down-regulation. As a result, we identified 25 B cell drug targets against RA and 23 B cell drug targets against SLE. We validated these targets by comparing them against the current treatment options for these diseases. For RA, five of the identified targets are genes that are already targeted by existing drugs used to treat RA. Our top-ranked targets for RA are adenine phosphoribosyltransferase, galactosidase beta 1, and aldehyde dehydrogenase 5 family member A1. For SLE, the top-ranked new targets are geranylgeranyl diphosphate synthase 1, glutathione-disulfide reductase, and galactosidase beta 1. Our pipeline provides a streamlined approach for identifying drug targets and predicting repurposable drugs and can facilitate the investigation of novel indications for diseases to improve global health. Supported by Layman Seed grant from University of Nebraska Foundation, and NIH grant R35GM119770.