NiChart^Workflows

A toolkit that allows users to apply NiChart processing methods and pre-trained models on their data. NiChart^Workflows includes tools for data harmonization and machine learning, using pre-trained models and reference distributions provided by NiChart^Engine. NiChart^Workflow utilizes Snakemake workflows to define multi-step image processing and data analytics pipelines.

Installation:

Install dependencies for the NiChart^Workflows in a conda environment using the following commands. NiChart^Workflows uses snakemake to automate things, which is (most easily) installed with mamba.

Note: Installation tested on macOS and Linux. Windows compatibility is not guaranteed.

conda install -n base -c conda-forge mamba # If needed
mamba init # If first time installing, then potentially source ~/.bashrc
mamba create -c conda-forge -c bioconda -n NiChart_Workflows python=3.8 
mamba activate NiChart_Workflows
pip install NiChart-Viewer-Demo spare-scores NiChartHarmonize
mamba install -c conda-forge -c bioconda snakemake

Note: Mamba installation from conda-forge may be slow. Linux x86 64 users can use a faster alternative:

1. wget https://github.com/conda-forge/miniforge/releases/latest/download/Mambaforge-Linux-x86_64.sh
2. bash Mambaforge-Linux-x86_64.sh
3. For installation, Yes everything
4. Re-start your kernel (re-open)
5. Check installation by typing "mamba" in the prompt

Usage:

Users can apply NiChart workflows to their data with a few simple steps:

1. Copy Data: Transfer your data to the designated input folder within the project directory.

2. Edit Configuration File: Make any necessary adjustments to the configuration file as specified by the package documentation.

3. Run Workflow: Execute the Snakemake workflow to initiate the analysis.

! For successful workflow execution, user data should meet specific requirements. Current version uses DLMUSE ROI volumes and sample demographic data as input. Please see the test dataset for an example.

! DLMUSE ROI volumes can be calculated by installing and running the DLMUSE package, or by applying it on the NiChart cloud

Examples:

We provided a test dataset "input/vTest1" as an example. The dataset includes demographic data (Study1_Demog.csv) and regional brain volumes (Study1_DLMUSE.csv) for a test set. Please use the test data as a reference for your input data.

You can apply the structural MRI workflow on this dataset as follows:

cd workflows/w_sMRI
snakemake -np           ## Dry run
snakemake --cores 1     ## Run the pipeline on Study1 data provided in the vTest1 dataset

If the workflow works successfully, workflow steps will be run on the data. Result files will be created inside the output folder:

"../../output/vTest1",

and the viewer will be launched with the combined result file as input.

Contributing:

We welcome contributions from the community! If you have bug fixes, improvements, or new features, please consider creating a pull request. Before submitting a pull request, please:

• Ensure your code adheres to the existing code style and formatting.
• Include clear documentation for your changes.
• Write unit tests for any new functionality.

License:

This project is licensed under the [License Name] license. Please refer to the LICENSE file for the full license text. (Replace [License Name] with the actual license used by your project, such as MIT, Apache, or BSD)

Contact:

For any inquiries, please contact guray.erus@pennmedicine.upenn.edu. (Last Updated: 5/24/2024)