[QUOTE=Frylock]
Am I right that if a human were to develop from the genetic code created by joining together the two half-codes contained in two randomly picked specimens of my sperm, then about 3/4 of that offspring’s genes would have two identical alleles?
My reasoning is:
I have either, say, DD, Dd, dD, or dd, each with the same probability
If I have DD, then my offspring will have DD.
If I have dd, then my offspring will have dd.
If I have Dd, then my offspring could have DD, Dd, dD, or dd
Same if I have dD.
Count up all the possible DDs, and you get 6/16. Count up the dds, and you get 6/16. Meaning 12/16, meaning 3/4.
This is different (right?) than the normal course of events, where the presence of a second contributor of DNA changes the probability so that just 1/2 of the genes consist of the same allele twice over.
-FrL-
[/QUOTE]
Umm…
The probabilities you give are true only if each gene has exactly two alleles and both are equally common in the population. This is a relatively rare situation. I don’t offhand remember the average number of alleles per gene, but I believe that the vast majority of genes have one allele which is overwhelmingly common (greater than 99% of people will have two copies) and a few alleles which are vanishingly rare, while a minority of genes have multiple alleles (in some cases hundreds) which vary substantially in frequency.
So for most of the offspring’s genes there would be two identical alleles in both cases, and for a minority of genes the uniparental offspring will have about a 1/2 chance of having two identical alleles (since the parent will most likely have two different alleles at these genes), while the normal bi-parental person will have a low probability (less than 1/2) of having two identical alleles at those genes.
This goes only for genes; non-gene DNA (which is most of the human genome) is much more variable.
JRB