PROJECT SUMMARY/ ABSTRACT Advances in neurotechnologies are producing large and complex datasets at unprecedented rate. Large-scale electrophysiological and optical imaging recordings provide an urging need to develop novel theoretical and analytical approaches to analyze and interpret these multi-scale, multi-resolution brain recordings. In response to the call of BRAIN Initiative 2.0 report, our project will develop new analytic tools and computational framework to understand complex and dynamic nature of large-scale spatiotemporal brain activity and the associated brain functions in a brain state-dependent manner. These tools will establish an analysis pipeline for large-scale electrophysiological and calcium imaging data. We will develop innovative supervised and unsupervised machine learning methods, and disseminate these analytic tools and software to the neuroscience community. Our team is uniquely positioned to not only develop these novel tools for basic neuroscience investigations, but also apply them in intracranial ECoG recordings where epilepsy patients underwent spatial or non-spatial memory tasks. In Aim 1, we will develop analytic tools and software for decoding representations of spatial or task information in large-scale hippocampal and neocortical recordings. In Aim 2, we will develop computational theories and framework for testing the task dimensionality of hippocampal population codes. In Aim 3, we will develop analytic methods and software for assessing and interpreting concurrent spatiotemporal neural patterns between multi-region, multi-scale, multi-resolution brain recordings. Overall, seamless integration of data analytics, theory and modeling as well as applications of these tools to address important basic/clinical neuroscience questions are highly significant. Accomplishment of these aims will not only establish an analysis pipeline for large-scale electrophysiological and calcium imaging data, but also empower experimental neuroscientists in their hypothesis-driven investigative studies and drive future data collection.