Graph AI generates mechanistic hypotheses validated across neurological systems

Neurological diseases are the leading global cause of disability, yet most lack disease-modifying treatments. We developed CIPHER, a graph AI model that generates mechanistic hypotheses for neurological disease. CIPHER uses a heterogeneous graph transformer contextualized to the adult human brain. CIPHER generated predictions across Parkinson’s disease (PD), bipolar disorder (BD), and Alzheimer’s disease (AD), which we validated using three independent biological systems. In PD, CIPHER linked genetic risk loci to genes essential for dopaminergic neuron survival and identified pesticides toxic to patient-derived neurons, including the insecticide Naled ranked within the top 6.75% of predictions. In silico CIPHER screens reproduced six genome-wide α-synuclein experiments, including a split-ubiquitin yeast two-hybrid system (normalized enrichment score [NES] = 2.27, FDR-adjusted p < 1E-4), an ascorbate peroxidase proximity labeling assay (NES = 2.22, FDR < 1E-4), and a high-depth targeted exome screen in 496 synucleinopathy patients (NES = 1.73, FDR < 1.9E-3). In BD, CIPHER nominated calcitriol as a candidate drug that reversed proteomic alterations in cortical organoids derived from BD patients. In AD, we conducted emulated clinical trials on cohorts involving n = 610,524 patients at Mass General Brigham, confirming that five CIPHER-predicted drugs were associated with reduced seven-year dementia risk (minimum hazard ratio = 0.63, 95% CI: 0.53–0.75, p < 1E-7). CIPHER generated and validated mechanistic hypotheses across molecular, organoid, and clinical systems, defining a path for AI-driven discovery in neurological disease.

Bio:
Ayush Noori (www.ayushnoori.com) is a Rhodes Scholar, Encode: AI for Science PhD Fellow, and D.Phil. student in Engineering Science in the Computational Health Informatics Lab at the University of Oxford. Driven by personal experiences, Ayush conducts research at the interface of artificial intelligence (AI), translational neuroscience, and precision medicine. He seeks to develop AI technologies that expand the frontier of personalized diagnosis and treatment for neurological disorders and other challenging medical conditions. His research efforts across Harvard Medical School, the Wyss Institute, and Massachusetts General Hospital have produced over 30 papers (including nine first or co-first author works) published in Cell, Nature Neuroscience, Nature Machine Intelligence, Nature Aging, npj Digital Medicine, Alzheimer’s & Dementia, and other peer-reviewed journals. Ayush received a Bachelor’s in Computer Science and Neuroscience and a concurrent Master’s in Computer Science from Harvard University.