Y-SCORCH 2.0: Further Data Mining and Functional Characterization for Single Cell Opioid Responses in the Context of HIV (SCORCH) Program

Yale University, New Haven CT

Investigators

Abstract

Opioid use disorder (OUD) and HIV, which affect 53 million and 40 million people respectively, exhibit a syndemic relationship due to structural social and health conditions that impact at least 20 million individuals globally. Both conditions contribute to central nervous system (CNS) dysfunction through distinct and synergistic pathways. Despite extensive research, the specific cellular mechanisms altered by OUD, HIV, and their interaction remain poorly understood. A comprehensive understanding of brain cell states and their localized responses is essential for developing targeted treatment strategies. The SCORCH initiative has facilitated unprecedented data generation through single-nucleus interrogation of brain tissue relevant to HIV and substance use disorders, providing an unparalleled opportunity to explore the molecular underpinnings of HIV, OUD, and their intersection. As the first data generation project supported by the SCORCH initiative, our Y-SCORCH program has significantly contributed to data generation and intellectual leadership. Y-SCORCH 2.0 aims to leverage our SCORCH experience and expertise to apply advanced tools to analyze transcriptomic and epigenetic data from the SCORCH consortium and novel spatial transcriptomic data to identify treatment targets. Y-SCORCH has already generated 346 single-nucleus transcriptomic and epigenetic datasets from the prefrontal cortex, insular cortex, and ventral striatum, across control, HIV, OUD, and HIV+OUD conditions. Through differential expression (DE) analysis, significant changes in immune activation and response have been identified, revealing widespread transcriptomic changes across cell types and elucidating cellular circuits involved in pathogenesis. The proposed research aims to mine SCORCH Consortium data to create a comprehensive catalog of DE genes in HIV and OUD across various brain regions and cell types. This will be achieved by applying standard protocols for DE analysis to ten brain regions and classifying DE genes in HIV versus controls, OUD versus controls, and HIV+OUD versus OUD cohorts. The team will also use neural network models to differentiate batch effects, intrinsic cell type representations, and disease-dependent variations. Additionally, the study aims to utilize neural networks and machine learning models to identify spatial signatures of HIV, OUD, and their interaction. Methods such as SIMVI (an unsupervised Variational Autoencoder) and SORBET (a supervised graph-neural network) will be used to analyze spatial transcriptomics data, identifying cell-cell interactions and expression gradients. The final aim is to leverage external datasets to further characterize the DE genes catalog and elucidate underlying mechanisms and etiology, using genome-wide association studies and other external resources. Y-SCORCH 2.0 will continue to advance the understanding of HIV, OUD, and their synergistic effects by employing cutting-edge methodologies, ultimately contributing to identification of therapeutic targets for these conditions.