A probabilistic framework for task-aligned intra- and inter-area neural manifold estimation

Low dimensional neural manifolds provide a compact characterization of population activity and shared co-variability across brain areas. Nonetheless technical limitations restrict the applicability of these tools to experimental designs without explicit trial structure, and thus cannot be used to understand neural responses during naturalistic behavior. Here we propose a novel probabilistic framework that allows for interpretable partitioning of population variability within and across areas in the context of naturalistic behavior. Our approach for task aligned manifold estimation (TAME-GP) explicitly partitions variability into private (to an area) and shared (across areas) sources, which can each be aligned to task-relevant axes. Furthermore, it uses biologically-appropriate Poisson noise, and introduces temporal smoothing of latent trajectories in the form of a Gaussian Process prior. This TAME-GP graphical model allows for robust estimation of task-relevant variability in local population responses, and of shared co-variability between brain areas. We demonstrate the efficiency of our estimator on simulated data. We also apply it to several datasets of neural population recordings in multiple species, areas and behavioral tasks. Overall, our results demonstrate the capacity of TAME-GP to capture meaningful intra- and inter-area neural variability with single trial resolution.