Project Overview

HIT-AI aims to transform drug repurposing for autoimmune disease from a serendipitous process into a scalable, predictive scientific methodology. Using knowledge graphs and machine learning applied to all 4,000 FDA-approved drugs across 640,000 drug-disease combinations, the center generates predictive scores for therapeutic matches, then validates high-potential candidates through high-throughput cellular screening, animal models, retrospective clinical studies using EHR data from 10M+ Americans, and clinical trials where possible. By targeting key autoimmune pathways — including TLR, TREX1-cGAS-STING, JAK-STAT, NF-κB, TCR, BCR, and BAFF — and integrating multi-omic data from Colton network members, HIT-AI aims to deliver at least three validated treatments to patients within three years.

Impact & Innovation

Making drug repurposing for autoimmunity predictable and scalable.

 

By mapping 640,000 drug-disease combinations with AI and validating hits through high-throughput screening and real-world clinical data, HIT-AI transforms the traditionally serendipitous process of drug repurposing into a systematic, cross-disease discovery engine.

  • Maps shared pathogenic pathways across autoimmune conditions at unprecedented scale, potentially explaining why drugs like rituximab and adalimumab work across multiple diseases and revealing new combination therapy opportunities
  • Generates IP and commercialization potential through novel dosing, formulation, and combination discoveries, with licensing, spinout, and industry collaboration strategies in development including partnership with Every Cure
  • Advances the Consortium’s Integrated Data and Discovery Platforms pillar by building a replicable, AI-driven repurposing infrastructure that leverages EHR data from 10M+ Americans and existing Colton clinical trial infrastructure across institutions
Research Approach

A framework designed for discovery

HIT-AI integrates AI-driven drug-disease prediction, high-throughput experimental screening, and real-world clinical validation to systematically identify and advance repurposed FDA-approved drugs for autoimmune diseases. The work combines computational knowledge graph analysis with cellular, animal, and clinical validation pipelines across multiple autoimmune conditions.

Knowledge graph construction and machine learning modeling to generate predictive scores for 640,000 drug-disease combinations across 4,000 FDA-approved drugs and 160+ autoimmune diseases; high-throughput cellular screening targeting key autoimmune pathways (TLR, TREX1-cGAS-STING, JAK-STAT, NF-κB, TCR, BCR, BAFF); multi-omic data integration from Colton network members; animal model validation of promising candidates; and retrospective clinical studies using EHR data from 10M+ Americans.

Biomedical knowledge graphs integrating drug, disease, and pathway data; multi-omic datasets from Colton network investigators; high-throughput cellular screening datasets targeting key autoimmune signaling pathways; EHR data from 10M+ Americans for retrospective clinical validation; and existing clinical trial infrastructure at UPenn and other Colton Centers.


Identification and clinical validation of repurposed drug candidates for 160+ autoimmune diseases, with a goal of delivering at least three validated treatments to patients within three years through laboratory validation, observational research, and clinical trials. Commercialization strategies include licensing, spinout companies, internal development at Penn, and collaboration with Every Cure, with IP anticipated from novel dosing, formulation, and combination therapy discoveries.

Investigators & Institutions

Powering the science

Principal Investigators

Jonathan Miner, MD, PhD, Colton Consortium Member

Associate Professor, Department of Medicine (Rheumatology), Perelman School of Medicine, University of Pennsylvania

Sara Cherry, PhD, Colton Consortium Member

John W. Eckman Professor of Medical Science, Department of Pathology and Laboratory Medicine, Perelman School of Medicine, University of Pennsylvania

David Fajgenbaum, MD, MBA, MSc, FCPP, Colton Consortium Member

Associate Professor, Department of Medicine (Translational Medicine and Human Genetics), Perelman School of Medicine, University of Pennsylvania

Research Outputs

From insight to impact

Publications

Serotonin reduction in post-acute sequelae of viral infection

Cell
Wong, AC; Devason, AS; Umana, IC; Cox, TO; Dohnalová, L; Litichevskiy, L; Perla, J; Lundgren, P; Etwebi, Z; Izzo, LT; Kim, J; Tetlak, M; Descamps, HC; Park, SL; Wisser, S; McKnight, AD; Pardy, RD; Kim, J; Blank, N; Patel, S; Thum, K; Mason, S; Beltra, JC; Michieletto, MF; Ngiow, SF; Miller, BM; Liou, MJ; Madhu, B; Dmitrieva-Posocco, O; Huber, AS; Hewins, P; Petucci, C; Chu, CP; Baraniecki-Zwil, G; Giron, LB; Baxter, AE; Greenplate, AR; Kearns, C; Montone, K; Litzky, LA; Feldman, M; Henao-Mejia, J; Striepen, B; Ramage, H; Jurado, KA; Wellen, KE; O'Doherty, U; Abdel-Mohsen, M; Landay, AL; Keshavarzian, A; Henrich, TJ; Deeks, SG; Peluso, MJ; Meyer, NJ; Wherry, EJ; Abramoff, BA; Cherry, S; Thaiss, CA; Levy, M October 2023
Animal ModelsBiological & MechanisticBiomarker DiscoveryCytokine SignalingExperimental Platforms & ModelsHuman CohortsNeuro-Immune InteractionsTranslational & ClinicalOtherUniversity of Pennsylvania

Expert perspective: diagnosis and treatment of Castleman disease

Arthritis & Rheumatology
Chen, LYC; Zhang, L; Fajgenbaum, DC June 2025
Biomarker DiscoveryDisease SubtypingDrug RepurposingExperimental Platforms & ModelsHuman CohortsTherapeutic DevelopmentTranslational & ClinicalCross-Cutting & Special PopulationsOtherRare Autoimmune DiseasesUniversity of Pennsylvania

ArfGAP2 promotes STING proton channel activity, cytokine transit, and autoinflammation

Cell
Poddar, S; Chauvin, SD; Archer, CH; Qian, W; Castillo-Badillo, JA; Yin, X; Disbennett, WM; Miner, CA; Holley, JA; Naismith, TV; Stinson, WA; Wei, X; Ning, Y; Fu, J; Ochoa, TA; Surve, N; Zaver, SA; Wodzanowski, KA; Balka, KR; Venkatraman, R; Liu, C; Rome, K; Bailis, W; Shiba, Y; Cherry, S; Shin, S; Semenkovich, CF; De Nardo, D; Yoh, S; Roberson, EDO; Chanda, SK; Kast, DJ; Miner, JJ March 2025
Animal ModelsBiological & MechanisticCytokine SignalingExperimental Platforms & ModelsFunctional Genomics & CRISPRHuman GeneticsIn Vitro ModelsInnate ImmunityAutoinflammatory DiseasesRare Autoimmune DiseasesUniversity of Pennsylvania

Crizanlizumab for retinal vasculopathy with cerebral leukoencephalopathy in a phase II clinical study

The Journal of Clinical Investigation
Wang, WX; Spiegelman, D; Rao, PK; Rhee, RL; Ford, AL; Miner, JJ; Apte, RS December 2024
Biological & MechanisticClinical TrialsDrug RepurposingExperimental Platforms & ModelsHuman CohortsHuman GeneticsInnate ImmunityPrecision MedicineTherapeutic DevelopmentTranslational & ClinicalCross-Cutting & Special PopulationsOtherRare Autoimmune DiseasesUniversity of Pennsylvania