The Data Science Core will be responsible for coordinating data management and analysis strategies across the five projects. Based at the University of Washington, the Core will integrate data science and computational neuroscience expertise to adapt or develop data analysis methods that can be applied in a coordinated way across the different data sets and to simulation outputs. Along with characterizing neural representations, we will particularly focus on new and developing methods that can extract information from small numbers of trials during learning, and that elucidate the dynamical basis of the neural computation. This analysis will be used in order to test and further develop hypotheses. The Core will establish standards and practices for data and software generated by and used the projects that will facilitate collaboration and ultimately sharing of data, algorithms and results. To facilitate efficient computation on large amounts of heterogenous data, the Data Science Core will develop portable high-performance computing systems that can be deployed on institutional resources, as well as using cloud computing platforms.
