The Development of a Novel Method for the Classification of the aCGH Profiles Based on Genomic Alterations

Due to recent advances in array technologies, array-based comparative genomic hybridization (aCGH) is now widely used to detect DNA copy number aberrations in cells in the biomedical research field, especially in the field of oncology. However, it may be difficult to make reproducible and consistent categorizations of aCGH profiles. The development of a convenient and reliable classification method for the aCGH profiles is therefore necessary for the clinical application of the aCGH technology. We recently developed a novel classification method for the a CGH profiles besed on genomic alterations in a reproducible fashion. This method has two steps, namely adaptive-weights-smoothing (AWS) and then a self-organizing map (SOM), which thus allows for the automatic classification of the aCGH profiles based on similarities in the genomic aberration pattern. We applied this method to 32 colorectal adenocarcinomas to demonstrate its practical utility. This method may thus be useful for identifying the genomic aberrations implicated in cancer phenotypes and clinicopthological correlations, because this method can be used for all aCGH data.

array CGH

colorectal cancer

classification

AWS (Adaptive Weights Smoothing)

SOM (Self Organizing Map)