A heuristic method of searching for interesting markers in terms of quantitative traits

Abstract

Selection of parental lines is important in plant breeding programmes. Marker-assisted selection is an alternative to classical selection methods, which are expensive and time consuming. Marker-assisted selection aims to find molecular markers that are linked to genes that determine quantitative traits of interest. Classical statistical methods require particular assumptions to be fulfilled, which is difficult to check if the analyses are performed automatically. In this article, we present a heuristic method to find interesting markers for quantitative traits. This method includes various strategies that depend on what makes a genotype interesting to a plant breeder. This approach was applied to eighteen parental lines of winter oilseed rape F1 CMS ogura hybrids with observation of 597 markers. The traits of interest were seed yield and alkenyl glucosinolate content. Fifty-seven markers were selected for further study. The most prominent marker was OPY 02~1830. Marker-assisted selection is the first step of analysis, which can then be followed up by a more formal statistical analysis for a smaller set of interesting markers.