caImageAnalysis

Introduction

Calcium imaging functional analysis

The default class is a BaseFish from fishy.py. We offer a process.py for various analyses. The standard implementation runs motion correction via caiman and then source extraction via suite2p, and most of the more advanced classes assume a suite2p folder with contained sources.

We leverage a folder structure that contains all experiment items

Folder Structure

Installation

This assumes you are using Anaconda and Python 3:

1. Install Mambaforge

2. Create a new environment and install Caiman. Note: helpful information on caiman can be found here

   mamba create -n caiman -c conda-forge caiman

3. Activate the envs (note: python should be 3.7x)

   mamba activate caiman

4. Install suite2p

   python -m pip install suite2p

5. Edit suite2p documentation to fit with scipy and numpy versions compatible with caiman

   \Lib\site-packages\suite2p\detection\sparsedetect.py

   Remove the 'keepdims' arg from line 256

6. Install more packages

   python -m pip install tables nptdms pyarrow

7. Optional: Download mimic_alpha into your new caiman envs - this package converts a list of RGB color that mimic a RGBA on a given background. This is useful for plotting.

Structure

core: motion correction, source extraction, calculate factors

utils: ideally things used multiple places, converts timestamps, stimuli, etc into dataframes

visualize: various visualization of response-classes, barcodes, neuron functions, etc

Issues: sometimes processing multiple fish at a time gives permission errors: unclear reproducibility, just refresh and give it another go