Contents - Index

GCA and SCA of parents - diallel analysis

From Biplot Tools, click Mean vs. Stability. The following will appear on the biplot:

• a small red circle indicating the position of the average tester;
• a thick red line that passes through the biplot origin and the average tester, referred to as the average-tester axis (ATA);
• a red arrow pointing to the average tester from the biplot origin;
• a thick blue line that passes through the biplot origin and is perpendicular to the average-tester axis; 
• a set of parallel lines that project each entry to the ATA; and
• two blue arrows on the thick blue line, pointing outwards from the biplot origin.

Interpretations on GCA:

• The entries are ranked based on their GCA along the ATA with the arrow pointing to a greater GCA value;
• The blue line separates entries with above-average GCA (on the same side of the 'average tester' relative to the biplot origin) from those with above-average PCA (on the opposite of the 'average tester');

Caution: The length of the average tester vector (the distance from biplot origin and the average tester), relative to the biplot size, is a measure of the relative importance of the entry main effect (GCA) vs. entry by tester interaction. The longer this distance, the more important is the GCA, and the more meaningful the ranking of parents based on GCA. At the extreme, a zero average tester vector means there is no GCA and therefore selection based on GCA is meaningless.

Interpretations on SCA:

• A longer projection to the thick blue line represents a greater positive SCA of an entry (in blue) with the testers (in red) on the same side of the ATA.
• A longer projection to the thick blue line represents a greater negative SCA of an entry (in blue) with the testers (in red) on the opposite side of the ATA.

Caution: When the biplot explains only a small portion of the total variation, the interpretation of the SCA may not accurate because some of the SCA are not displayed.

Identifying heterotic groups:

• Contrasting groups of entries relative to the ATA may be regarded as different heterotic groups.

  Note:

• If the biplot explains only a small proportion of the total variation, this interpretation may not be accurate for short-vector parents. 
• In a diallel cross dataset, each genotype is both an entry and a tester. GCA and SCA are properties of "entries", rather than "testers".
• However, the testers can also be evaluated for their ability to identify entries with high GCA, in analogous to the test enivironment evaluation in genotype by environment data analysis.