BBQS AI Resource and Data Coordinating Center (BARD.CC)

SUMMARY Understanding the complex relationship between brain activity and behavior is one of the most exciting and challenging pursuits in neuroscience. The proposed BBQS AI Resource and Data Coordinating Center (BARD.CC) aims to facilitate innovative research in this area by managing, sharing, and harnessing the power of vast amounts of data and machine learning resources generated by various projects within the BBQS consortium. We will focus on five interrelated aims: 1) Data Management; 2) Data Standards; 3) Machine Learning and Artificial Intelligence (ML/AI) Resources; 4) Data Ecosystem; and 5) Dissemination, Training, and Coordination. The first aim is to serve as a hub for efficient data curation, management, and sharing. We will collaborate with other BBQS projects and coordinate with existing BRAIN data archives to curate and harmonize project data. Data management will be handled by a combination of automated data ingestion and human oversight, transitioning to a fully automated system over time. We will work with scientists and relevant communities to implement robust quality assurance and control solutions. The second aim focuses on establishing data standards for novel sensors and multimodal data integration, as informed by the use of existing standards and best practices from similar efforts. We will aggregate relevant standards for data and metadata, data processing methods, appropriate ontologies, and common data elements, and adapt as needed for evolving methodologies. The third aim involves the development and definition of ML/AI resources for BBQS. We will evaluate and curate relevant ML/AI models and platforms, aggregating datasets, models, and other ML/AI resources from both within and outside the BBQS consortium. These resources will be made available to consortium members, with each resource's origin documented and evaluated for performance and ethical generation and use. Moreover, all models will be made available through public repositories, allowing for widespread access and utilization. The fourth aim involves creating a cloud-based data ecosystem and computational platform. We will collaborate with relevant archives and computing facilities to develop a computational platform in the cloud. This platform will enable access to and processing of even very large data sets with commonly used pipelines and provide a wide range of users, even those with limited resources, with computational capability to analyze and visualize data, models, and model outputs. Finally, the fifth aim is centered around efficient dissemination, training, and coordination of BBQS research resources. We will coordinate data sharing, offer training on relevant topics like neuroinformatics, neuroethics, and ML/AI, and maintain a consortium Web portal. Furthermore, center staff will coordinate consortium activities like meetings, working groups, and policy and ethics discussions, ensuring smooth and effective operation. In summary, BARD.CC aims to catalyze the discovery of valuable insights from intricate relationships between brain activity and behavior, which in turn could advance neuroscience and our understanding of the human mind.