bryanlimy / OASIS

Deconvolution of calcium imaging data

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

OASIS: Fast online deconvolution of calcium imaging data

Tools for extracting the neural activity from fluorescence calcium imaging data  

The code can be readily run on neural temporal fluorescence calcium imaging data. Please have a look at the demo.

Requirements

The scripts were tested on Linux and MacOS (some users successfully used Windows too) with a typical numerical/scientific Python 2.7 or 3.5-3.7 installation, e.g. using Anaconda or Canopy, that included the following

  • python >= 2.7.11
  • matplotlib >= 1.5.1
  • numpy >= 1.10.2
  • scipy >= 0.16.1
  • cython >= 0.23.4

Optionally, because not necessary for running our fast method on your own data, we further installed the following to perform the comparison with interior point methods

Installation

For faster execution some functions have been written in Cython and need to be compiled by running: python setup.py build_ext --inplace
(Ignore the warnings that Cython is using a deprecated Numpy API. Following the cython online docs, for the time being, it is just a warning that you can ignore.)
(If the compilation fails on Windows, see Srikanth's gist.)

To clean up temporary files follow it by: python setup.py clean --all

Examples

The scripts to produce the figures and table are in the subfolder 'examples' with names obvious from the PLoS Comput Biol paper. They can be run with ipython examples/fig[1-6].py.

The results of fig4 and table1 will be even better than in the paper, because the version in the master branch includes later improvements, in patricularly up to an order of magnitude less computing time. The specific points in history marking the time of the publications have been tagged.

To demonstrate how to use the methods on your own data, we included a demo jupyter notebook in the subfolder 'examples' as well.

Other implementations

Related packages

In order to deal not just with temporal, but with raw spatio-temporal fluorescence data, we added OASIS also to CaImAn, the computational toolbox for large scale Calcium Imaging Analysis.

References

The code accompanies a short NIPS paper and an extended journal version with full details. If you use our code in your research, please cite one of them:

About

Deconvolution of calcium imaging data

License:GNU General Public License v3.0


Languages

Language:Python 100.0%