Systematic evaluation of variability in ChIP-chip experiments using predefined DNA targets

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Chromatin immunoprecipitation on tiling microarrays (ChIP-chip) is a key method for genome-wide protein-DNA interaction detection.
• Objective analysis of ChIP-chip platforms, protocols, and algorithms is crucial for reliable results.

Purpose of the Study:

• To objectively analyze the performance of different tiling array platforms, amplification procedures, and signal detection algorithms in ChIP-chip experiments.
• To establish a benchmark for evaluating future ChIP-chip technologies and methods.

Main Methods:

• A simulated ChIP-chip experiment using human genomic DNA and controlled "spike-ins" at various concentrations.
• Hybridization to four tiling array platforms by eight independent groups performing blind predictions of spike-in locations.

Main Results:

• Microarray platform choice was not the primary determinant of performance; inter-lab and inter-protocol variations were greater.
• Long oligonucleotide arrays showed slightly higher sensitivity for low enrichment.
• Simple sequence repeats and genome redundancy caused false positives across platforms.
• Linear-Molding PCR (LM-PCR) and Whole Genome Amplification (WGA) accurately reproduced relative enrichment levels.
• Signal detection algorithm performance varied significantly depending on the array platform.

Conclusions:

• Standardization of laboratory protocols and analysis methods is critical for improving ChIP-chip reliability.
• Array platform characteristics influence cost-effectiveness and detection power, especially concerning tiling resolution and replicates.
• The developed spike-in samples and data provide a valuable resource for future ChIP-chip research and development.