pyComBat, a Python tool for batch effects correction in high-throughput molecular data using empirical Bayes methods.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Technical biases, known as batch effects, introduce variability into gene expression datasets beyond biological variation.
• ComBat and ComBat-Seq are established tools for correcting batch effects in microarray and RNA-Seq data, respectively.

Purpose of the Study:

• To introduce a novel Python implementation of the ComBat and ComBat-Seq algorithms.
• To evaluate the performance and utility of this new implementation for the bioinformatics community.

Main Methods:

• Developed a Python package named pyComBat, available as part of the inmoose package.
• Maintained the original mathematical framework of ComBat and ComBat-Seq.
• Ensured open-source distribution under GPL-3.0 license with publicly accessible source code.

Main Results:

• The pyComBat implementation achieves comparable batch effect correction accuracy to the original R versions.
• The Python implementation demonstrates equal or superior computational speed compared to the original tools.
• The new implementation facilitates community involvement in further development.

Conclusions:

• A new Python version of ComBat and ComBat-Seq (pyComBat) has been developed for efficient batch effect correction.
• This implementation retains the efficacy of the original tools while offering enhanced performance and accessibility.
• pyComBat provides a valuable resource for researchers working with microarray and RNA-Seq data.