CHIPIN: ChIP-seq inter-sample normalization based on signal invariance across transcriptionally constant genes
Lélia Polit1, Gwenneg Kerdivel1, Sebastian Gregoricchio2
1Institut Cochin, Inserm U1016, CNRS UMR 8104, Paris Descartes University UMR-S1016, 75014, Paris, France.
BMC Bioinformatics
|August 18, 2021
Summary
This study introduces CHIPIN, an R package for normalizing ChIP-seq data without spike-ins, using gene expression levels. CHIPIN improves accuracy for analyzing protein binding and chromatin structure across conditions.
Area of Science:
- Genomics
- Molecular Biology
- Bioinformatics
Background:
- ChIP-seq experiments are crucial for studying gene modulation and drug effects on protein binding and chromatin.
- Current normalization methods for ChIP-seq signals often rely on flawed assumptions, potentially leading to inaccurate biological interpretations.
Purpose of the Study:
- To develop a novel method for normalizing ChIP-seq signals across different conditions when spike-in controls are unavailable.
- To provide a user-friendly tool that enhances the reliability of ChIP-seq data analysis.
Main Methods:
- Developed the CHIPIN R package for ChIP-seq signal normalization.
- Utilized gene expression data as a basis for normalization, assuming no significant ChIP-seq signal differences in regulatory regions of consistently expressed genes.
- The method is also applicable to ATAC-seq and DNase hypersensitivity data.
Main Results:
- CHIPIN enables accurate ChIP-seq normalization without spike-in controls, leveraging gene expression data.
- The package offers visualization tools and statistics to evaluate normalization efficiency and antibody specificity.
- CHIPIN demonstrated superior performance compared to existing normalization techniques in validation studies.
Conclusions:
- CHIPIN offers a robust and accessible solution for ChIP-seq signal normalization in the absence of spike-in experiments.
- The R package is publicly available on GitHub, facilitating its adoption in the research community.
