Contents - Index


QTL identification - multiple traits in a single environment

 

This function requires a genotype by (marker+trait) two-way table. The analysis is similar to generating a covariable-effect biplot except that the explanatory variables are genetic markers and the response variables are various traits. 

 

The analysis involves the following steps:

 

 

The identification of QTL for 6 traits are illustrated below (achieving each of the biplots is just a mouse-click away):

 

1) The original marker by trait biplot with all markers:

 

 

2) Marker by trait biplot when markers with short vectors are removed:

 

 

3) Marker by trait biplot in which each possible QTL is represented by two linked markers:

 

 

4) Marker by trait biplot in which each possible QTL is represented by a single marker:

 

 

These biplots may be referred to as QQT biplot, in contrast to the QQE biplot for barley yield

 

NOTE: This biplot is more meaningful when the target traits are somehow inter-related.

 

In Depth: 

Why the No-Centering option is used in the QQT or QQE biplotThis is because the values in the r-matrix are "effects" of the traits/markers on the target trait in each environment, and the purpose of the the QQT or QQE biplot is to visualize these effects. Any centering will distort the effects and any biplot on distorted data will be misleading.