Federal Grant Award Summary

Cold Spring Harbor Laboratory (CSHL) received a $460,629 Project Grant from the National Institute on Drug Abuse (NIDA) under the Drug Use and Addiction Research Programs (CFDA 93.279) effective March 15, 2026, with a completion date of December 31, 2028. The project develops and delivers advanced machine learning methodologies—specifically knockout training algorithms—designed to identify precise one-to-one mappings between artificial neural network model units and biological brain neurons. These algorithms enable neuroscientists to incorporate and interpret neural perturbation data by training models on behavioral observations from animals with experimentally inactivated neurons, thereby elucidating how individual neurons represent sensory information and generate motor behaviors.

The deliverables include improved knockout training algorithms optimized for datasets with limited training data, neural recording integration, and high behavioral variability; connectome-informed computational models of the Drosophila melanogaster visual system leveraging recently released connectomic anatomical data; and comprehensive validation of the proposed algorithms using both simulated and real perturbation data from fruit fly neural systems. CSHL will make these general-purpose knockout training algorithms widely available to the systems neuroscience research community, supporting the broader national research objectives in understanding neural mechanisms relevant to substance use and addiction research priorities.