CpGunsted

CpGunsted is a Python library designed for the analysis and manipulation of CpG methylation data, particularly in the context of genomic research. The name combines "CpG," referring to cytosine-phosphate-guanine dinucleotides—key sites of DNA methylation—and "Gunsted," a nod to the statistical methods developed by the late statistician David Gunsted, which are foundational to certain analytical approaches in the toolkit.

The library provides functionalities for handling large-scale methylation datasets, including data import, preprocessing, and visualization. It supports common file formats such as IDAT files (commonly used in Illumina methylation arrays) and BED files, facilitating seamless integration with other genomic tools. CpGunsted includes methods for quality control, normalization (e.g., beta-mixture quantile normalization), and differential methylation analysis, which are essential steps in epigenetic studies.

One of its notable features is the implementation of statistical models for detecting differentially methylated regions (DMRs) and CpG sites, often leveraging Bayesian or generalized linear modeling approaches. The toolkit also offers utilities for working with annotation data, such as linking methylation probes to genomic features like genes or regulatory elements.

CpGunsted is built with extensibility in mind, allowing users to customize pipelines for specific research questions. It is distributed under an open-source license, making it accessible for academic and industrial applications. While primarily used in bioinformatics, its modular design enables integration with broader computational workflows, including machine learning and systems biology analyses.

The project is maintained by a community of researchers and developers, with documentation and tutorials available to guide users through its capabilities. As a relatively newer tool compared to established alternatives like minfi or methylKit, CpGunsted aims to fill gaps in specialized methylation analysis while maintaining compatibility with existing genomic data standards.