A flexible spot recognition method for SNP microarray systems

Huang CY, Liu L

Abstract

Image based Microarray processing has recently found widespread application in biotechnology. With the dramatic increase in the number of genotyping assays, microarray systems can be utilized in a high throughput setting to analyze large numbers of samples and SNPs quickly and efficiently. The flexibility on different number of SNPs and sample size becomes more and more important to accommodate different types of research. However, the key in employing this technology successfully is the ability to accurately detect the spotted samples during image processing on different image layout. In this paper, we propose a highly flexible and automated method to acquire spot intensity and status to achieve this goal. Since there could be many different combinations on the microarray layout, the image processing method was designed with two distinct processes; a rough and precise detection using a Circular Template image bipartition, and mesh algorithms that can automatically and accurately process spot layouts with minimal information required. For a high throughput system, automatically detecting the spots with their intensity and classifying the status of spots are important tasks for automated genotype calling. The methodology presented in this paper can automatically locate spots on an assay plate and reports their Foreground Intensity, Background Intensity and status. The approach described in this paper can be applied on plate, slide, or other type of container for both gene-expression and SNP genotyping image based system. The overall accuracy of spot detection was 99.98%. The total processing time is fast enough to generate more than 1 million genotyping assays per day to be a high-throughput system.