29 - Whole Genome SNP Array Improves Diagnosis and Therapy in Pediatric Brain Tumors

Abstract

can be extremely useful information when proceeding with a full chromosomal copy number analysis of a sample. As the sample is evaluated, absolute copy number (versus relative copy number) is preferred for final reporting. The problem in general is extremely difficult if not impossible using data just from SNP arrays or NGS. An algorithm to estimate percent aberrant cells from a normal population was devised using B-allele frequency estimations derived from either SNP array or next generation sequencing results. This approach includes an automated calculation per copy number event for the percentage of cells that are aberrant to help distinguish mosaicism among the full sample population. Additionally, along with the type of copy number alteration (i.e. gain, loss), an estimate for absolute copy number will be reported for each event, if such estimation is possible. This tool was tested on a series of cancer samples with known aberrations. Estimation of percent aberrant cells and absolute copy number was possible for most regions of copy number alteration. Exceptions were limited to highly complex samples with mixed mosaicism.