Quantized correlation coefficient for measuring reproducibility of ChIP-chip data.

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Chromatin immunoprecipitation followed by microarray hybridization (ChIP-chip) is crucial for studying genome-wide protein-DNA interactions and histone modifications.
• Replicate experiments in ChIP-chip are essential for data quality assessment, typically using correlation coefficients to measure reproducibility.
• Standard correlation coefficients can be misleading as they are influenced by both signal reproducibility and the amount of binding signal present.

Purpose of the Study:

• To develop a novel statistical measure for assessing ChIP-chip data reproducibility.
• To create a correlation coefficient that is less dependent on the magnitude of the binding signal.
• To improve the accuracy and robustness of reproducibility assessment in genomic experiments.

Main Methods:

• Developed the Quantized Correlation Coefficient (QCC) by discretizing data into quantiles.
• Implemented a merging procedure to group background probes and reduce noise influence.
• Recalculated the Pearson correlation coefficient on the processed data.

Main Results:

• QCC demonstrates significantly less dependence on signal amount compared to standard correlation coefficients.
• QCC accurately reflects reproducibility across ChIP-chip replicates with varying signal enrichment and coverage.
• The developed method is more robust than Pearson or Spearman correlation coefficients for ChIP-chip data analysis.

Conclusions:

• QCC provides a more reliable measure of ChIP-chip data reproducibility, especially when signal levels vary.
• The quantization and merging procedures in QCC can aid in distinguishing signal from background noise.
• QCC is applicable beyond ChIP-chip, offering a robust method for assessing reproducibility in array CGH and gene expression data.