SNPknock documentation¶
Introduction¶
SNPknock is a Python library for generating knockoff variables from discrete Markov chains and hidden Markov models, with specific support for genomic data. See also the corresponding R package.
This package implements the algorithms for knockoff generation described in:
- “Gene hunting with hidden Markov model knockoffs”, Sesia et al., Biometrika, 2019, (doi:10.1093/biomet/asy033).
- “Multi-resolution localization of causal variants across the genome”, Sesia et al., bioRxiv, 2019, (doi:10.1101/631390).
Feature highlights:
- Generate knockoffs for discrete Markov chains (DMC).
- Generate knockoffs for hidden Markov models (HMM).
- Generate knockoffs for genotype and haplotype data.
- Provides a user-friendly interface for fitting an HMM to genetic data using the software fastPhase.
If you want to learn about applying SNPknock to analyze data from large genome-wide association studies, see KnockoffZoom: https://msesia.github.io/knockoffzoom.
SNPknock is licensed under the GPL-v3 License.
Source code¶
The source code is hosted on GitHub: https://github.com/msesia/snpknock-python/