Robust Predictors of Mania and Psychosis

点击翻译按钮获取中文摘要

ADMINISTRATIVE SUPPLEMENT PROJECT SUMMARY: This application is an administrative supplement in response to PA-20-272 for a year of extension with funds. The major goals of the supplement are to 1) make data collected and tools developed under the parent grant available to the wider scientific community, and 2) extend data collection in a subset of participants whose data are deemed highly valuable. In Aim 1, we propose to curate 30 multimodal longitudinal (1-year long) datasets to the common format and ethical standard of the National Institute of Mental Health data archive (NDA), including de-identification and transcription of clinical interviews, for sharing with interested investigators in the Intensive Longitudinal Health Behaviors Network (ILHBN) and the wider scientific commnunity. We will also work to disseminate the Deep Phenotyping ToolKit developed under the parent grant, which includes tools for the aggregation and quality control, interactive visualization, and processing of multimodal longitudinal data. In Aim 2, we propose to study 10 select participants for longer than the one-year period designated in the parent grant. The participants will be selected based on having temporally dense multimodal data and fluctuating symptoms. To formalize the process of participant selection, we will develop a tool for forecasting the value of a participant’s longitudinal data, based on data missingness and variance in clinical scores in an initial study period. While the Covid pandemic has had an impact on the parent project, this supplement proposal goes far beyond simply compensating for adverse effects of the pandemic on productivity. The extended goals of the supplement rely on the integration of data and accumulation of knowledge afforded by the parent grant, and could not have been accomplished within the scope of the parent grant. Overall, making multimodal longitudinal datasets, and tools to visualize and process multimodal longitudinal data, available across and beyond the ILHBN community, will maximally leverage the potential of these assets for a broad array of investigators nationally and globally, and will foster the development of common standards for handling longitudinal data. Wider access to high-quality, temporally-dense, behavioral and clinical data in individuals with a mental illness will facilitate and accelerate rigorous computational phenotyping research of severe mental states.