A stable density approach to probe selection for a custom aCGH design

Usage of the custom aCGH design provides the ability to detect copy number alternations (CNAs) with a very high resolution in almost every genomic region. Since the detection rate of CNAs greatly depends on the quality of the design, it is crucial to ensure the optimal probe coverage for the regions of interests. In this paper, we focus on the problem of finding the best possible probe coverage for a given region and a predefined set of probes. The lack of available probes at some places is a considerable problem which may lead to the decrease in the detection rate of CNAs. We propose a new approach towards the generation of coverage with stable probe density. The developed algorithms attempt to compensate the lack of probes in poorly covered regions by selecting additional probes from their nearest neighborhood. Here we introduce evaluation measures that reflect density stabilization along the covered region. Finally, we use these measures to compare our algorithms with other standard approaches for coverage preparation.