A shifting level model algorithm that identifies aberrations in array-CGH data

作者：Magi Alberto^*; Benelli Matteo; Marseglia Giuseppina; Nannetti Genni; Scordo Maria Rosaria; Torricelli Francesca

来源：Biostatistics, 2010, 11(2): 265-280.

DOI：10.1093/biostatistics/kxp051

• array-CGH
• Segmentation algorithm
• Shifting level model

Array comparative genomic hybridization (aCGH) is a microarray technology that allows one to detect and map genomic alterations. The goal of aCGH analysis is to identify the boundaries of the regions where the number of DNA copies changes (breakpoint identification) and then to label each region as loss, neutral, or gain (calling). In this paper, we introduce a new algorithm, based on the shifting level model (SLM), with the aim of locating regions with different means of the log(2) ratio in genomic profiles obtained from aCGH data. We combine the SLM algorithm with the CGHcall calling procedure and compare their performances with 5 state-of-the-art methods. When dealing with synthetic data, our method outperforms the other 5 algorithms in detecting the change in the number of DNA copies in the most challenging situations. For real aCGH data, SLM is able to locate all the cytogenetically mapped aberrations giving a smaller number of false-positive breakpoints than the compared methods. The application of the SLM algorithm is not limited to aCGH data. Our approach can also be used for the analysis of several emerging experimental strategies such as high-resolution tiling array.

• 出版日期2010-4