Item

Abstract

To evaluate the suitability of massive parallel sequencing by Illu- mina/Solexa sequencing technology for variant detection and allele frequency estimation, we sequenced the MUTYH gene in two pools of DNA (from breast cancer patients). A 6 kb long-range PCR (LRP) was designed containing exons 2-16. One pool consisted of 287 ge- nomic DNA samples, serving as template for LRP. The second pool consisted of 88 LRP-products derived from individual samples. Equi- molarity of the constituent samples was calculated from concentration measurements with fluorimetry for genomic DNA and high resolution melting curve analysis (HR-MCA) for LRP-products. Illumina sequenc- ing results were compared to Sanger sequencing results of individual samples. Correlation between allele frequencies detected by both methods seemed poor in the first pool, probably due to variable DNA quality among samples, a too large pool size and unequal amplification caused by an Alu insertion polymorphism. Frequencies correlated well in the second pool, which allowed reliable detection of a frequency of 2 in 176 alleles (1.1%) or higher, whereas 2 of the 5 singletons detected by Sanger were significantly above background noise in the Illumina output. These results provide directions in designing high-throughput analyses of candidate genes in large series of patients and controls.