Memorial Sloan-Kettering Cancer Center has received a USD 1.12 million NIH supplemental award to build machine learning models of gene regulatory networks in rheumatoid arthritis, using T cells and synovial fibroblasts as the primary cellular targets.
The grant, issued through the National Human Genome Research Institute's IGVF (Impacts of Genomic Variation on Function) program, funds work led by Alexander Y. Rudensky and Christina S. Leslie. The project addresses a core challenge in polygenic disease genetics: most disease-associated variants lie in non-coding regulatory regions, making it difficult to connect genetic variation to cellular phenotypes across multiple interacting cell types.
The team will train allele-specific gene regulatory models integrating single-cell multiomic data with bulk 3D chromatin interaction maps, using genetically diverse F1 hybrid mice as a training resource before applying transfer learning to human RA patient samples. Parallelized Perturb-seq experiments in primary synovial fibroblasts from RA patients will be used to validate and refine the models. Spatial transcriptomics on inflamed joint tissue will additionally map interactions between T cell and fibroblast populations within local tissue microenvironments.