PyMethylProcess
https://github.com/Christensen-Lab-Dartmouth/PyMethylProcess
Wiki:
https://github.com/Christensen-Lab-Dartmouth/PyMethylProcess/wiki
Help documentation: https://christensen-lab-dartmouth.github.io/PyMethylProcess/
Alternatively, you can access the pdf: PyMethylProcess.pdf
What is it:
- Preprocess 450k and 850k methylation IDAT files in parallel using Minfi, ENmix, and meffil
- Convenient and scalable implementation
- Imputation and Feature Selection
- Preparation for machine learning pipelines
Why:
- Make DNAm accessible to python developers and more machine learning oriented researchers
- Streamlined analysis makes processing easy
PyMethyProcess is now available in Bioinformatics: https://academic.oup.com/bioinformatics/advance-article/doi/10.1093/bioinformatics/btz594/5542385 .
Getting Started:
- Installation:
- Example Usage Scripts (in github repo): Located in ./example_scripts/
- Help docs (in github repo): https://christensen-lab-dartmouth.github.io/PyMethylProcess/
- See GitHub wiki for more information on getting started, preprocessing quickly, and setting up and running machine learning analyses / classic methylation analyses (cell type deconvolution, age estimation).
- Running the CWL tool (temporary instructions until new docker upload):
- Clone this repository.
- sh docker_build.sh
- Then execute the cwl/pymethylprocess.cwl tool using Toil https://toil.readthedocs.io/en/latest/ or Rabix Composer or executor https://github.com/rabix/composer, amongst others.
- Try this dataset for quick testing: GSE109541
- Note: This CWL tool has limited functionality, if you would like to see additional functions automated (eg. age calculation, cell type deconvolution, running machine learning pipelines), please submit an issue, and we'll add new features.
NOTE: There have been reported issues with installing PyMethylProcess on Mac OS Mojave (rpy2). If this is the issue, try the docker installation and please report an issue.
CWL Workflow Visualization
Benchmark Results:
Supplementary Figure Removed from Manuscript:
Supplemental Figure 1: UMAP embeddings (colored) of: a) GSE87571 (age), b) GSE81961 (disease status), c) GSE69138 (subtype), d) GSE42861 (disease status), e) GSE112179 (brain disorder), f) GSE90496 (subclass), g) TCGA Pancancer (subtype)