Accurate identification of communication between multiple interacting neural populations

TL;DR

This paper introduces MR-LFADS, a novel model that accurately disentangles inter-regional neural communication, unobserved inputs, and local dynamics, outperforming existing methods in simulations and real data analysis.

Contribution

The paper presents MR-LFADS, a new sequential variational autoencoder that improves identification of neural communication between brain regions, addressing limitations of previous models.

Findings

MR-LFADS outperforms existing methods in simulated multi-region networks.

It accurately predicts effects of circuit perturbations in large-scale neural data.

The model effectively disentangles inter-regional communication and local dynamics.

Abstract

Neural recording technologies now enable simultaneous recording of population activity across many brain regions, motivating the development of data-driven models of communication between brain regions. However, existing models can struggle to disentangle the sources that influence recorded neural populations, leading to inaccurate portraits of inter-regional communication. Here, we introduce Multi-Region Latent Factor Analysis via Dynamical Systems (MR-LFADS), a sequential variational autoencoder designed to disentangle inter-regional communication, inputs from unobserved regions, and local neural population dynamics. We show that MR-LFADS outperforms existing approaches at identifying communication across dozens of simulations of task-trained multi-region networks. When applied to large-scale electrophysiology, MR-LFADS predicts brain-wide effects of circuit perturbations that were held out during model fitting. These validations on synthetic and real neural data position MR-LFADS as a promising tool for discovering principles of brain-wide information processing.