plinkformatter 0.1.44

No project description provided

PLINKFORMMATER

This repository is designed to transform genotype data from the Muster SNPs/download endpoint into PLINK-compatible data, which can then be processed by PyLMM. The primary aim is to facilitate genomic data transformation for linear mixed models. This tool should, in theory, also work with GEMMA and other software that consumes standard PLINK file formats, though it has been primarily tested with PyLMM.

Getting Started

Prerequisites

To use this repository, you must have the following dependencies installed:

• Python 3.8 or above
• PLINK 2.0
• Poetry

Install dependencies:

poetry install

Activate virtual environment:

poetry shell

Install PLINK:

You can download PLINK from the official website. Ensure the PLINK executable is in your system's PATH, or you can specify the path to the PLINK binary in your environment settings.

To verify PLINK installation, run:

plink2 --version

Tests

To run tests run the following command:

pytest -s tests

Publishing to Pypi

0. Update version

poetry version patch

1. Build any changes

poetry build

2. Set the correct PyPI repository URL

poetry config repositories.pypi https://upload.pypi.org/legacy/

3. Set API token

poetry config pypi-token.pypi pypi-YourActualTokenHere

4. Publish

poetry publish

TODO

Differences from Hao's R code:

• Pheno ordering matters too
• Massive pheno file difference. Need to de-deuplicate
• stop upper casing strains, Hao mentioned this.
• Hao creates kinship matrix from pedmap not bfile
• Hao uses --pedmap PLINK flag not --ped --map, switch in an effort to keep everything the same.
• confirm that the first column of the .pheno file is zscore: expected strain | strain | zscore | value -> matches actual
• generate_pheno_plink_fast still uses create_sorted_pgen, Hao's code does not leverage the pfiles. -- [x] sanity check, ensure no p-files created 45911_plink_new_v1 has only bed bim bam like Hao.
• we have an autosome only filter on --chr 1-19