Analysis of genomic abnormalities in tumors: a review of available methods for Illumina two-color SNP genotyping and evaluation of performance

Abstract

Several methods have recently been proposed for identifying copy number alterations (CNAs) in genomic DNA from tumors, using the signals arising from two-color genotyping technologies. Although copy number estimation in normal tissue has been well studied, methods developed for normal tissue tend to perform poorly when applied to tumors, due to normal cell contamination, varying levels of ploidy, and genetic heterogeneity within the tumor. Here we compare the performance of seven methods (DNA-Chip Analyzer software (dCHIP), GenoCNA software, allele-specific copy number analysis of tumors (ASCAT), OncoSNP software, genome alteration print (GAP) visualization, CNVpartition software plug-in for the Genome Studio software, and Partek Genomics Suite software) that have been established for two-color CNA analysis on the Illumina platform, using two ovarian cancer cell lines where spectral karyotyping analysis has also been performed, and two tissue samples, one from a highly malignant ovarian cancer and one from a benign ovarian tumor, all of which harbor significantly different genomic abnormalities. ASCAT shows very stable estimates of CNAs, as does OncoSNP when jointly analyzing paired normal DNA. We found the best performance, in general to be from ASCAT.