Contents - Index

QTL identification - single trait in multiple environments


Note: This module requires a genotype-by-(marker+environment) two-way table


QTL identification based on phenotypic data from multiple environments is one example of covariate effect analysis; another example is the analysis of a genotype by (trait+environment) two-way table.


This analysis involves the following steps, which are automated in GGEbiplot:


In depth:

This QTL mapping process involves two steps:

1) Individual marker screening (point statistics)

2) Global threshold (LR for QTL regression and vector length for QQE) screening



Large effect QTL represented by very closely linked markers can be missed if the LR is used as a global error control method, because the markers would have very small LR. An extreme example would be two markers at almost the same spot. Remove one marker would result in little reductions in the residual SS and therefore a very small LR.  

When using the vector length as a criterion in QQE analysis (i.e., identifying QTL data on data from multiple environments), one should be aware that a large number of closely linked markers tend to have longer vectors whereas unlinked markers tend to have shorter vectors, relative to their true effects. However, this is less of a problem than QTL identification based on data from a single environment.