PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Microarray image segmentation using additional dye - An experimental study
Rashi Gupta, Salla Ruosaari, Sangita Kulathinal, Jaakko Hollmen and Petri Auvinen
Molecular and Cellular Probes Volume 21, Number 5-6, pp. 321-328, 2007.

Abstract

The DNA microarray technique allows monitoring the expression levels of thousands of genes simultaneously. A single DNA microarray experiment involves a number of error-prone manual and automated processes, which influence the results and have an impact on the subsequent stages of analysis. Typical problems of arrays are pinning errors while probe printing and the corruption of spots by noise patches. These errors should be detected at the time of image analysis in order to prevent the erroneous intensities from ending up in the analysis and inference stages. Results: In this paper we introduce the concept (referred to as SybrSpot) of utilizing information provided by an additional dye, SYBR green RNA II, for segmentation of gene expression microarrays. Owing to the effective binding of the SYBR green RNA II to the array probes, an image with high signal-to-noise ratio is obtained. This image is used to learn about the spot quality and to flag spots which are not reliably hybridized and corrupted by noise. Further, we compare SybrSpot with GenePix and demonstrate that SybrSpot performs better than GenePix when flagging spots with no probes or weak probes.

EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
ID Code:3609
Deposited By:Jaakko Hollmen
Deposited On:13 February 2008