Estimation and assessment of raw copy numbers at the single locus level

摘要: Motivation: Although copy-number aberrations are known to contribute the diversity of human DNA and cause various diseases, many their phenotypes still be explored. The recent development single-nucleotide polymorphism (SNP) arrays provides researchers with tools for calling genotypes identifying chromosomal at an order-of-magnitude greater resolution than possible a few years ago. fundamental problem in array-based (CN) analysis is obtain CN estimates single-locus high accuracy precision such that downstream segmentation methods more likely succeed. Results: We propose preprocessing method estimating raw CNs from Affymetrix SNP arrays. Its core utilizes multichip probe-level model analogous high-density oligonucleotide expression extend this by adding adjustment sequence-specific allelic imbalances as cross-hybridization between allele A B probes. focus on total estimates, which allows us further constrain increase signal-to-noise ratio estimates. Further improvement obtained controlling PCR effects. Each part fitted robustly. performance assessed quantifying how well alone differentiate one two copies Chromosome X (ChrX) (27kb) up 200kb resolution. evaluation done publicly available HapMap data. Availability: proposed open-source R package named aroma.affymetrix. Because it bounded-memory algorithm, any number can analyzed. Contact: hb@stat.berkeley.edu Supplementary information: Supplementary data Bioinformatics online.