x-linked recessive inheritance and correlation for males

Question

I was inspired by a discussion in this thread. Wikipedia lists a number of disorders linked to recessive genes on the x-chromosome. One typical example is red-green color blindness. Now wikipedia says that up to ten percent of men and up to one percent of women suffer from red-green color blindness. This is due to the fact that the recessive gene for red-green color-blindness sits on the x-chromosome and men have only one copy of this chromosome while women have two.

Now if you take a random male that suffers from red-green color blindness there is a ten percent chance that his father also has red-green color blindness. There is also a ten percent chance that his mother has red-green colorblindness. She must have one copy for red-green color blindness, and if so there is a ten percent chance that she is homozygote and thus suffers from color blindness herself.

So if you did not know the gene causing red-green color blindness and investigated, say 1000 red-green color blind males you will find no correlation between their color blindness and the sex of a parent that also have color blindness.

If you do it the other way around you will of course find that 100 percent of the sons of color blind women be color blind, there is total correlation. So you get very different results depending on the way you set up your investigation.

Now there are also polygenetic traits such as length or intelligence that might be affected by recessive genes on the x-chromosome. The y-chromosome is much shorter than the x-chromosome and contains fewer genes than the x-chromosomes so you would naively expect that somewhat more of the variance in polygenetic traits in male are due to genes inherited from their mother than from their father.

Now if you set up a large scale investigation trying to find out how much of the variation in a polygenic trait in a male that is due to genes from his father and from his mother will the contribution from x-linked recessive genes be accurately determined?

If a man has an x-linked recessive gene making him, say, one centimeter taller than if he had another gene variant, there is equal likelyhood that his father and his mother is affected by this gene perhaps making the correlation "invisible"?

If a man does not have a recessive x-linked gene there is zero likelihood that his mother is affected by this gene but his father could still be affected by that gene.

Question: 1. How does the fact that some of the x-linked genes contributing to a polygenic trait are recessive influence large scale investigations based on correlations trying to determine how much of the variation in a polygenic trait in a male that is due to genes inherited from his mother?

Answer 1

The correlations you are looking are all based on some unusual and arbitrary subset of the data. The one type of correlation that people generally look at is the one that we call heritability.

For a detailed definition of the term heritability, please have a look at this post. In short, heritability (in the narrow sense; again see above link) is the fraction of the overall phenotypic variance (for a specific trait) that is caused by additive genetic variance.