A single-array preprocessing method for estimating full-resolution raw copy numbers from all Affymetrix genotyping...

Area of Science:

• Genomics
• Bioinformatics
• Statistical Genetics

Background:

• High-resolution copy-number (CN) analysis is crucial for identifying CN aberrations and polymorphisms.
• Optimizing statistical methods is essential for cost-effective and successful CN studies.
• Existing methods require careful consideration of array types and data preprocessing.

Purpose of the Study:

• To propose a single-array preprocessing method for estimating full-resolution total CNs.
• To develop a method applicable to all Affymetrix genotyping arrays, including those with non-polymorphic probes.
• To optimize statistical methods for accurate and efficient CN analysis.

Main Methods:

• A single-array preprocessing method (CRMA v2) for estimating total CNs.
• The method controls for allelic crosstalk, probe affinities, PCR fragment-length effects, probe sequence effects, and co-hybridization.
• A reference signal is only needed for calculating relative CNs.

Main Results:

• CRMA v2 performs comparably to or better than existing methods (Affymetrix CN5, dChip) in differentiating CN states.
• The method achieves full-resolution analysis with varying degrees of smoothing.
• CRMA v2 demonstrates the feasibility of online analysis for large-scale projects with evolving datasets.

Conclusions:

• The proposed single-array method provides accurate and efficient high-resolution CN analysis.
• CRMA v2 is a valuable tool for genomic studies investigating CN aberrations and polymorphisms.
• The method supports scalable and adaptable analysis pipelines for large genomic projects.