Project Grant R35GM166095

The National Institute of General Medical Sciences awarded The Ohio State University $383,169 on July 23, 2026, under the Biomedical Research and Research Training program (CFDA 93.859) to develop AI-powered computational models for translational research that benchmark the fidelity of animal models and organoid systems against human references and construct ensemble virtual cells to predict human cellular responses under perturbations.

The award funds two integrated research directions. Direction 1 establishes a fidelity framework that applies variance decomposition, batch reasoning, and heterogeneous graph transformers with contrastive learning to rigorously compare organoid and animal models against human biological data, disentangling biological signals from technical noise and generating fidelity scorecards quantifying human relevance across regulatory programs, cell-cell communication, and tissue architecture. Direction 2 constructs ensemble virtual cells that integrate existing gene-expression, protein, and regulatory models into a universal representation using hypergraph transformers to harmonize molecular, cellular, and tissue-scale data and predict human-like cellular responses to perturbations.

The work addresses the translational gap where fewer than ten percent of therapies effective in mice ultimately succeed in humans, and where current virtual cell technologies remain fragmented, lack mechanistic grounding, and produce inconsistent perturbation predictions.

The award period extends through June 30, 2031, with place of performance at The Ohio State University in Columbus, Ohio.