Artificial Intelligence and Machine Learning Core

Project Summary: The Artificial Intelligence and Machine Learning (AI/ML) Core (AIC) will significantly expand upon the existing Bioinformatics and Pathways Core (BPC) to provide COBRE Research Project Leaders (RPLs) with AI/ML expertise and support. This expansion is crucial for RPLs to leverage AI/ML in their research, particularly in the context of high-throughput data analysis, and to remain competitive in securing R01-level grants. AI/ML, especially transformer technology (e.g., ChatGPT), is rapidly transforming various sectors of society, including biomedical research. The AIC will assist COBRE RPLs, Pilot Project Awardees, and the wider genomic regulation research community in Oklahoma in adopting these technologies. The core will offer advanced skills, expertise, and specialized computational infrastructure (GPUs) to better facilitate the adoption of state-of-the-art AI/ML tools in research. Furthermore, we will engage in the development and customization of new genomic analysis tools to understand complex genomic regulation. The AIC will provide expertise and resources for AI/ML tasks like image analysis and transcriptional network analysis, including developing classifiers for cytology images, assisting in data analysis and integration, and denoising imaging data. The AIC will develop a genomic region embedding system for genomic interval analysis, which will use deep learning to capture spatial relationships and features within DNA sequences, going beyond traditional statistical approaches to evaluate regions of interest (ROI). It will allow for the integration of various data types (multi-omics) and help infer regulatory implications of region sets. It will be an important component of helping interpret the significance of newly gathered genomic ROI in the context of prior experiments, genomic region annotations, epigenetic alterations, significance of biological pathway enrichments, and physiochemical genomic properties.