Single-cell multi-omics of patients and AI-augmented bioinformatics to discover the signaling loop driving chronic disease progression.

We develop and apply AI-augmented computational models on the multi-omics data of diseased tissue to systematically discover the inter- and intra-cellular signaling pathways in the tissue and their alternations in the disease. These data include single-cell, bulk and spatial transcriptomics, spatial imaging, proteomics, interactome, etc. for integrated analysis and cross-validation. Together with our understanding of the disease's clinical phenotypes, we identify the key pathways that drive disease progression. From these pathways, we select novel targets for wet-lab validation, based on their importance in the disease as well as safety and modality considerations.

Various autologous cell types of patients co-cultured on a 3D multi-channels and stretchable chip to recreate the patient tissue microenvironment.

This platform integrates organoid and organ-on-a-chip technologies to overcome the low clinical translatability issue of animal efficacy models in drug R&D. Guided by our deep understanding of key cell types and signaling pathways in a disease, a variety of cell types, including but not limited to epithelial, endothelial, fibroblast and immune cells, are assembled together on demand into the Emulate chip and form a 3D tissue. The cells can be either primary cells or iPSC cells from patients and controls. The tissue in the chip can functionally mature with a precise control of a fluid flow and a mechanical stretch that mimics physiological phenomena in vivo. This in vitro human model recapitulates many physiological features of human tissue in vivo and may better validate drug targets in preclinical studies and predict drug efficacies in clinical trials.

Omics data-based animal disease model selection and customization to recapitulate the target signaling pathway in human patients.

Given the differences in species and cause of a disease phenotype, animal models cannot fully recapitulate the pathophysiology of a human disease. For a particular target in a disease, we compare the omics data of patients and the known animal models of the disease to select the animal model that best recapitulates the target signaling pathway in patients. If better pathway recapitulation is needed, we customize the animal model by adjusting its experimental protocol. After an in vivo study with compound treatment, we analyze animal samples to validate the perturbation of the target pathway. We leverage both internal and external in vivo platforms for pre-clinical target validation.