It depends on what you're measuring. If the pedigree is known, then you regress the phenotype on the relationship matrix. The extent to which you can predict the phenotype based on the relationship matrix will give you the genetic correlation.
If you have "unrelated" (but multiply tested) individuals, like in a controlled cross with sibs or clones, then do an ANOVA and see the proportion of variance explained by family, and that will be your genetic correlation.
From a "this is a textbook problem" perspective, you would plot the average of the scores of the mother and father on the x-axis and the individuals' scores on the y-axis, do this for all subjects, then find the slope of the line. That's the broadsense heritability, also known as mid-parent regression. And genetic correlation
What's your design? RCBD? how does your dataset looks like?
Dear Silvas
The estimation of variance components depend on the degree of relationship between relatives , mean the individuals included in the study , and the nature of data ( scale , ordinal , nominal )
Good luck
Dear Silvas
working by MATLAB software is very easier than SAS. I recommend you to write related program according to your experimental design in MATLAB. in addition its likely that there is no any way to calculate genetic correlation using SPSS.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS