Archiver > GENEALOGY-DNA > 2004-07 > 1089141441

From: Ruy Cardoso <>
Subject: Re: SNP discovery
Date: Tue, 6 Jul 2004 12:17:21 -0700 (PDT)
In-Reply-To: <>

> From:
> Date: Mon, 5 Jul 2004 14:22:03 EDT

> The descendant haplotypes do form a cluster, but not

> everyone with the same haplotype has an ancestor
> that SNP. [snip] His descendants begin to accumulate

> variations on that one haplotype, forming a
> star-shaped network diagram. [snip] For more recent
> haplogroups, say R1b3c and R1b3f, we're more likely
> to find the same haplotype occurring in two
> haplogroups.

Ann -

I found your explanation on this topic to be most
cogent. Naturally, it raises more questions.

(1) Your description seems to imply that the mutation
process for SNPs is independent of that for STRs. Is

I suppose the answer could be yes (statistically) and
no (biologically), but I'm guessing that, in practice,
statistical independence would preclude the ability to
see biological DEpendence.

(2) Assuming that the answer to (1) is yes, then is it
reasonable to say that haplogroup results are not all
that useful within a genealogical time frame (i.e.,
within, say, twenty generations)?

Perhaps haplogroup results could be useful for
eliminating spurious matches when only a small number
of STRs are tested, but I'm under the impression that
it would be more cost-effective to just test more
STRs. I am not suggesting, by the way, that
haplogroup results are uninteresting beyond the
genealogical time frame or that they won't be become
more interesting as more is learned.

(3) As testing large numbers of STRs becomes more
cost-effective, will the usefulness of SNPs decline?

Then again, perhaps the degree of variation for
haplotypes within haplogroups can provide additional
information about STR mutation rates.

Thanks for any additional comments.

Ruy Cardoso

Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!

This thread: