mesmerize-core

Mesmerize core backend

Installation | Examples

A batch management system for calcium imaging analysis using the CaImAn library. It contains pandas.DataFrame and pandas.Series extensions that interface with CaImAn for running the various algorithms and organizing input & output data.

This replaces the Mesmerize legacy desktop application.
mesmerize-core is MUCH faster, more efficient, and offers many more features! For example there are simple extensions which you can just call to get the motion correction shifts, CNMF reconstructed movie, CNMF residuals, contours etc.

Documentation

We recommend starting out with the demo notebook notebooks/mcorr_cnmf.ipynb

Documentation is available at: https://mesmerize-core.readthedocs.io/
User guide: https://mesmerize-core.readthedocs.io/en/latest/user_guide.html

Getting Help

Please use the GitHub issue tracker for any issues, and discussions for discussions. We no longer use gitter. If you are in the Slack, we usually only respond on slack around workshops. GitHub is for long term support.

Video tutorial/virtual workshop from September 2022: https://www.youtube.com/watch?v=0AGiAaslJdk

Overview

Visualization

We recommend either fastplotlib or mesmerize-viz for visualization. Mesmerize-viz gives you ready-to-use viuslizations for motion correction and CNMF - including component exploration, interactive component evaluation using metrics, and manual addition or removal of components.

Note that mesmerize-viz is very new and we are always happy to help (same for mesmerize-core and fastplotlib!). Installation instructions for mesmerize-viz are at the repo, it has very few dependencies so installation should be straightforward: https://github.com/kushalkolar/mesmerize-viz

:exclamation: Harware requirements The large CNMF visualizations with contours etc. usually require either a dedicated GPU or integrated GPU with access to at least 1GB of VRAM.

Motion correction visualization using mesmerize-viz

https://github.com/nel-lab/mesmerize-core/assets/9403332/34413fa4-6041-4418-ab66-ddd7dca03e19

Explore CNMF variants

https://github.com/nel-lab/mesmerize-core/assets/9403332/3453d6fb-0dae-4520-b2c2-8156d8ce7b58

CNMF Componment Evaluation

https://github.com/nel-lab/mesmerize-core/assets/9403332/b6e8f6df-6bd4-4167-96eb-38c5095c8770

If you use fastplotlib directly you can create highly complex and interactive plots to combine outputs from the CaImAn algorithms with other experimentally relevant analysis, such as behavioral data.

Examples

See the notebooks directory for detailed examples.

Note that fastplotlib is required for the visualizations.

Installation

For users

The instructions below will install mesmerize-core.

For visualization install fastplotlib into the same environment as mesmerize-core:

pip install "fastplotlib[notebook]"

You may need to install Vulkan drivers depending on your system, see the fastplotlib repo for more information: https://github.com/kushalkolar/fastplotlib#graphics-drivers

Conda

mesmerize-core is availabe as a conda package which also gives you CaImAn! These instructions will give you a working mesmerize-core along with caiman in the same environment.

Important note: Sometimes conda or mamba will get stuck at a step, such as creating an environment or installing a package. I found that pressing Enter on your keyboard can sometimes help it continue when it pauses.

1. Install mamba into your base environment. Skip this step if you have mamba. This step may take 10 minutes and display several messages like "Solving environment: failed with..." but it should eventually install mamba.

conda install -c conda-forge mamba

# if conda is behaving slow, this command can sometimes help
conda clean -a

2. To create a new environment and install mesmerize-core into it do this:

mamba create -n mescore -c conda-forge mesmerize-core

caiman is a dependency of mesmerize-core so it will automatically grab caiman too

If you already have an environment with caiman:

mamba install -n name-of-env-with-caiman mesmerize-core

3. Activate environment. You can only use mesmerize-core in the environment that it's installed into.

mamba activate mescore

4. Install caimanmanager

If you are using an older version of caiman < 1.9.16, then please see Step 4 in the old README.

caimanmanager install

If you have issues with this step, please post an issue on the caiman github or gitter: https://github.com/flatironinstitute/CaImAn/issues

5. Run ipython and verify that mesmerize_core is installed:

# run in ipython
import mesmerize_core
mesmerize_core.__version__

6. Install fastplotlib for visualization into the same environment (run this in the anaconda prompt, not ipython). You may also need to install Vulkan drivers depending on your system, see the fastplotlib repo for more information: https://github.com/kushalkolar/fastplotlib#graphics-drivers

pip install "fastplotlib[notebook]"

Optional: install simplejpeg for much faster notebook visualization, you will need C compilers and libjpeg-turbo for this to work:

pip install simplejpeg

python virtual environments

# create a new env in some directory
# tested on python3.9 and 3.10
python3.10 -m venv python-venvs/mesmerize-core
source python-venvs/mesmerize-core/bin/activate

# get latest pip setuptools and wheel
pip install --upgrade setuptools wheel pip

# cd into or make a dir that has your repos
mkdir repos
cd repos

# install caiman
git clone https://github.com/flatironinstitute/CaImAn.git
cd CaImAn
pip install -r requirements.txt
pip install .
caimanmanager install

# install mesmerize-core
pip install mesmerize-core

# install fastplotlib
pip install "fastplotlib[notebook]"

# you should now be able to import mesmerize_core
# start ipython
ipython

# run in ipython
import mesmerize_core
mesmerize_core.__version__

For development

conda

# install mamba in your base environment
conda install -c conda-forge mamba
conda clean -a

conda create --name mesmerize-core

# activate environment
conda activate mesmerize-core
conda clean -a

# clone this repo
git clone https://github.com/nel-lab/mesmerize-core.git
cd mesmerize-core

# update env with environment file
# this installs caiman as well
mamba env update -n mesmerize-core --file environment.yml

# install caimanmanager
caimanmanager install

# install mesmerize-core
pip install .

# install pytest and run tests to make sure everything works properly
mamba install pytest
MESMERIZE_KEEP_TEST_DATA=1 DOWNLOAD_GROUND_TRUTHS=1 pytest -s .

python venvs

# create a new env in some directory
# tested on python3.9 and 3.10
python3.10 -m venv python-venvs/mesmerize-core
source python-venvs/mesmerize-core/bin/activate

# get latest pip setuptools and wheel
pip install --upgrade setuptools wheel pip

# cd into or make a dir that has your repos
mkdir repos
cd repos

# install caiman
git clone https://github.com/flatironinstitute/CaImAn.git
cd CaImAn
pip install -r requirements.txt
pip install .
caimanmanager install

# clone this repo and install mesmerize-core
cd ..
git clone https://github.com/nel-lab/mesmerize-core.git
cd mesmerize-core
pip install -e .

# run tests to make sure everything works
MESMERIZE_KEEP_TEST_DATA=1 DOWNLOAD_GROUND_TRUTHS=1 pytest -s .