Similarity Measurement of Array-based Comparative Genomic Hybridization Data in Evaluation of Whole Genome Amplified DNA
*Ching-Wei Chang, NCTR, FDA  James Fuscoe, NCTR, FDA  Tao Han, NCTR, FDA 


Genome-wide genotyping assays often require substantial amounts of input DNA. To overcome the problem of limited amounts of clinical DNA samples, Whole Genome Amplified (WGA) methods have been developed. The uniformity of the amplification process has not been extensively characterized and array-based comparative genomic hybridization (aCGH) was used to assess the uniformity. Two WGA methods were evaluated by aCGH. Similarity measurements are necessary for detecting possible differences between amplified and non-amplified DNA, and between WGA methods. Several similarity measurements, including Pearson's correlation, Kolmogorov distance, and time series similarity measurement, will be discussed in this study. Smoothing algorithms are also an important issue in aCGH data analysis. How the smoothing algorithm affects the similarity measure will also be discussed.