Home Available Analyses FAQ Docs
Sign In

LocaNMF (Saxena et al. 2020)


Analysis description

Localized semi-Nonnegative Matrix Factorization (LocaNMF) is a method that efficiently decomposes widefield video data and allows us to directly compare activity across multiple mice by outputting mouse-specific localized functional regions that are significantly more interpretable than more traditional decomposition techniques. Moreover, it provides a natural subspace to directly compare correlation maps and neural dynamics across different behaviors, mice, and experimental conditions, and enables identification of task- and movement-related brain regions.

Useful links
LocaNMF (Saxena et al. 2020) Paper Link
LocaNMF (Saxena et al. 2020) Github Repo Link
LocaNMF (Saxena et al. 2020) Bash Script Link
LocaNMF (Saxena et al. 2020) Demo Link
How to use this analysis

The NeuroCAAS implementation of LocaNMF works with version 1.1 of LocaNMF. Look for live logging in log.txt as well as standard DATASET_NAME files.

Args:
  -Low-rank widefield video: (mat file) The .mat file called 'Vc.mat' should contain a variable 'U' of low-rank spatial components with dimensions [fov_width,fov_height,rank], a variable 'V' of low-rank temporal components with dimensions [rank,timepoints,trials] (note that number of trials can be 1), and a variable 'brainmask' of a mask denoting the brain in the field of view.
  -Config: (yaml) a yaml file containing the following parameters:
     1) names of the variables in the input data file ('Vc.mat'),
     2) maxrank: how many max components per brain region.
     3) min_pixels: minimum number of pixels in Allen map for it to be considered a brain region.
     4) loc_thresh: Localization threshold, i.e. percentage of area restricted to be inside the 'Allen boundary'.
     5) r2_thresh: Fraction of variance in the data to capture with LocaNMF.
     6) Path to aligned atlas (mat file): 'atlas.mat' should contain the Allen atlas, registered / aligned to the field of view. Check out https://github.com/ss5513/locaNMF-preprocess to preprocess your data. Example given in template file.

Outputs:
  -LocaNMF_Components (folder): The spatial and temporal components in variables 'A' and 'C' respectively, with the corresponding region names for each component in 'areanames'.
  -Figures (folder):
     CorrelationPlot.png: correlations between different LocaNMF regions;
     SummaryComponents.png: A summary of the different spatial and temporal components for every regions.


You must login to use an analysis.

  • Twitter
  • Github
  • Dribbble