MIAAIM: multi-omics image alignment and analysis by information manifolds

MIAAIM is a software to align multiple-omics tissue imaging data. The worflow includes high-dimensional image compression, registration, and transforming images to align in the same spatial domain. MIAAIM was developed at the Vaccine and Immunotherapy Center at MGH in the labs of Dr. Patrick Reeves and Dr. Ruxandra Sîrbulescu.

For further documentation on the MIAAIM Python impementation, please visit joshuahess12.github.io/miaaim-python.

Installation

You can install MIAAIM in Python using either the MIAAIM-Python Docker container, which would allow for complete workflow reproducibility, or you can install the package into your environment with pip.

Dependencies

MIAAIM utilizes the Elastix library for image registration computations, which is written in the C++ language. For this reason, we recommend running your workflows with the MIAAIM Python package inside of a Docker container, which we have created to automatically include Elastix. You can still run MIAAIM, however, if you would rather stick with installing packages via pip, you will just need to install Elastix separately. These two options for installing MIAAIM are outlined below:

Cloning the repository:

To clone the repository directly, use the following command to ensure that all submodules are included:

git clone https://github.com/JoshuaHess12/miaaim-python.git --recurse-submodules

Usage without Docker / Install with Pip:

If you are unable to use Docker on your machine, then you can still use MIAAIM:

1. download the latest version of Elastix.
2. Make Elastix accessible to your $PATH environment (Ex. on a Mac, access your .bash_profile and add export PATH=~/elastix-latest/bin:$PATH and export DYLD_LIBRARY_PATH=~/elastix-latest/lib:$DYLD_LIBRARY_PATH)
3. Run the following command to install MIAAIM on your machine:

 pip install miaaim-python  # install miaaim

Reproducibility with Pip

If you are using pip to install MIAAIM, you can reconstruct your working environment easily with the commands: 1.

 pip freeze > requirements.txt  # create documentation of installed packages

This will create a text file that indicates the specific packages that you are using. You can then install the specific packages that were exported into another environment with : 2.

pip install -r requirements.txt

We have included a requirements.txt file in this repository to use for our convenience.

Docker

MIAAIM's Python implementation is containerized using Docker to enable a reproducible environment. Inside of this container, the Python distribution of MIAAIM is already installed. It is therefore set up so that users can copy scripts and data into it in order to run analyses that they need.

To get started with MIAAIM using Docker:

1. Install Docker.
2. Ensure that Docker is available to your system using the command docker images
3. Pull the miaaim-python docker container docker pull joshuahess/miaaim-python:latest where latest is the version number.

Using MIAAIM inside of Docker

If you are using MIAAIM with Docker, we recommend having a concrete file structure for data and code with relative paths so that your script doesn't rely on absolute file paths outside of the Docker container.

You can mount your custom scripts and data into the virtual environment as follows: 4. Mount your data and scripts into Docker from your local path (src-path)

docker run -it -v /path/to/data:/data joshuahess/miaaim-python:latest bash    # mount data in the "dest-path" folder

Here, we assumed that the folder data contains your new script and your input data that goes with it. 5. Run your script (named here script.py) from the data folder:

python ./data/scipt.py

Note here that any additional packages that you use to process your data that are not included in the docker image will not be found!