Multiple loci in silico mapping in inbred lines

Abstract

The in silico mapping (ISM) technique and its extension represent major advances for novel gene discovery in germplasm resources of inbred lines. However, the techniques suffer from a relatively high false-positive rate (FPR) and they do not consider the effect of linkage disequilibrium (LD) markers around the identified quantitative trait locus (QTL). In addition, it has not yet been established whether it is optimal to use absolute trait differences as the response variable. To address these problems, this article presents the multiple loci ISM (MLISM) approach, which uses all markers on the entire genome, along with a penalized maximum likelihood. The method proposed here was verified by a series of simulation experiments with a maize pedigree population of inbred lines of known ancestry. Results from the simulated studies show that the best response variable is the trait product. The MLISM FPR is substantially decreased and the proportion of the number of false QTL to the number of LD markers around the identified QTL is adequately reduced. The MLISM method, with the trait product as the response variable, is an improvement on the existing methods for novel QTL mapping in germplasm resources of inbred lines.

Epistatic Association Mapping in Homozygous Crop Cultivars

“…First, main, environmental, QTL-by- environment and QTL-by-QTL interactions were simultaneously considered in our full genetic model, improving the statistical power [38]–[41]. Although multi-locus genetic models have been proposed in plant genetics [59]–[62], they have difficulty combining both QTL-by-environment and QTL-by-QTL interactions. Epistasis association mapping has been developed in human genetics [33]–[37], but here the epistasis was identified by two-dimensional scan, and significant effects in the two-dimensional scan were further tested in one genetic model.…”

Bias correction for estimated QTL effects using the penalized maximum likelihood method

“…Although the scaled inverse chi-square has been considered as a prior distribution for s j 2 in the PML method of Lü et al (2009), there are some subtle differences between that method and the one proposed here. The former seeks to decrease the false positive rate in the detection of QTL by increasing the value of t, whereas the latter seeks to enhance the QTL detection ability by decreasing the value of t. In this study, we selected a suitable value for t that maintained a high statistical power and a low false positive rate.…”