Is this the right normalization? A diagnostic tool for ChIP-seq normalization.

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• ChIP-seq is standard for genome-wide protein-DNA interaction profiling.
• Normalization is crucial for ChIP-seq data to correct biases between ChIP and Input DNA samples.
• Current assessment of normalization methods lacks comprehensive evaluation and diagnostic tools.

Purpose of the Study:

• To propose a novel diagnostic tool for evaluating ChIP-seq normalization appropriateness.
• To assess existing normalization methods using the proposed diagnostic tool.
• To introduce a new False Discovery Rate (FDR) control procedure using sample swapping.

Main Methods:

• Development of a diagnostic plot based on empirical densities of log relative risks.
• Evaluation of normalization estimates from CisGenome, NCIS, and CCAT using the diagnostic plot.
• Comparison of peak calling results using different normalization constants with tools like MACS and SICER.
• Implementation of a novel FDR control procedure via sample swapping.

Main Results:

• The proposed diagnostic plot effectively assesses the appropriateness of normalization constants.
• The choice of normalization constant significantly impacts peak calling tools (MACS, SICER).
• The novel FDR control procedure gains power by utilizing estimated normalization constants.

Conclusions:

• Linear normalization estimates scale factors to adjust for sequencing depth in ChIP-seq data.
• The diagnostic plot aids in selecting adequate ChIP/Input normalization constants for improved peak identification accuracy.
• Incorporating appropriate normalization constants enhances downstream ChIP-seq data analysis.