Latent Factor Analysis via Dynamical Systems


Neuroscience is experiencing a revolution in which simultaneous recording of thousands of neurons is revealing population dynamics that are not apparent from single-neuron responses. This structure is typically extracted from data averaged across many trials, but deeper understanding requires studying phenomena detected in single trials, which is challenging due to incomplete sampling of the neural population, trial-to-trial variability, and fluctuations in action potential timing. We introduce latent factor analysis via dynamical systems, a deep learning method to infer latent dynamics from single-trial neural spiking data. When applied to a variety of macaque and human motor cortical datasets, latent factor analysis via dynamical systems accurately predicts observed behavioral variables, extracts precise firing rate estimates of neural dynamics on single trials, infers perturbations to those dynamics that correlate with behavioral choices, and combines data from non-overlapping recording sessions spanning months to improve inference of underlying dynamics.
LFADS Architecture Figure




  • Guide to Multisession Modeling with lfads-torch - Topics:
    • Formatting data inputs to train a multisession lfads-torch model via neural stitching
    • Training a model on multisession data
    • Evaluating the resulting latent factors and firing rates inferred by lfads-torch

lfads-torch Discussion on Gitter

The public Gitter room provides an opportunity to discuss lfads-torch with the developers and follow pull requests, issues, and releases. To join the Gitter room, click the button below and sign up for an Element account via your GitHub credentials. You can also join via Matrix.

AutoLFADS on NeuroCAAS

NeuroCAAS is a cloud platform that connects users with Amazon Web Services cloud computing resources via a simple web interface. We have implemented AutoLFADS to run on NeuroCAAS with support for standard and multisession model training.


@article{Pandarinath_OShea_Collins_Jozefowicz_Stavisky_Kao_Trautmann_Kaufman_Ryu_Hochberg_et al._2018, 
  title={Inferring single-trial neural population dynamics using sequential auto-encoders}, 
  journal={Nature Methods}, 
  author={Pandarinath, Chethan and OShea, Daniel J. and Collins, Jasmine and Jozefowicz, Rafal and Stavisky, Sergey D. and Kao, Jonathan C. and Trautmann, Eric M. and Kaufman, Matthew T. and Ryu, Stephen I. and Hochberg, Leigh R. and et al.}, 
  title={A large-scale neural network training framework for generalized estimation of single-trial population dynamics}, 
  journal={Nature Methods}, 
  author={Keshtkaran, Mohammad Reza and Sedler, Andrew R. and Chowdhury, Raeed H. and Tandon, Raghav and Basrai, Diya and Nguyen, Sarah L. and Sohn, Hansem and Jazayeri, Mehrdad and Miller, Lee E. and Pandarinath, Chethan},