GENEALOGY-DNA-L ArchivesArchiver > GENEALOGY-DNA > 2006-02 > 1139868263
From: (John Chandler)
Subject: Re: [DNA] Earliest common ancestors for 37/37 and 43/43 matches
Date: Mon, 13 Feb 2006 17:04:23 -0500 (EST)
> database of samples large enough to "assure" ANY mutation will occur
> would have to be very large...certainly not just 74 samples.
First, it isn't necessary (or even possible) to assure that ANY
mutation will occur. Rather, it is just necessary to assure
(or render likely) that at least one mutation will occur TWICE.
Second, although mutations are random, they are not rare in
projects that cover multi-generation lineages. Look again at
the numerical example I gave a couple of days ago -- by assuming
an average depth of 10 generations to MRCA in a group of 25
people, I came up with an expectation of 21 mutations. The
prediction was then 96% of at least one pair of parallel
mutations. You reach 99% chance with a group of 31 people.
In some circles, that counts as assurance.
What you need to do is look carefully at a project that has
an apparent case of parallel mutation and see how many mutations
there are in the family group in question. The more you see, the
more likely there will be parallel mutations.
|Re: [DNA] Earliest common ancestors for 37/37 and 43/43 matches by (John Chandler)|