GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-05 > 1115222421


From:
Subject: Re: [DNA] Newbie: National Geographic, Relative Genetics, DNA
Date: Wed, 04 May 2005 16:00:21 +0000


Actually Ken, FTDNA is providing a valuable service by assigning a person to the broad I category since as you well know there are haplotypes that could just as easily be G, I or J - take a guess. Once a person knows that they are I, there are people such as yourself on this List who could "tilt the probability" toward I1a, I1b or I2 or even finer categorizations such as "Ultra Norse".

Would it be possible at this stage to construct an algorithm, as Whit has done for many of the general haplogroups, so people could plug their numbers into a tool and get the figures crunched on the spot? As long as people understand that since Y-STRs mutate at a relatively high rate there is an unknown error rate attached to the predictions all this seems reasonable. Perhaps you could recommend that anyone whose score falls beyond a preset confidence interval on the algorithm opt for SNP testing.

With all due respect Ken, your efforts need to move to the next stage, meaning the development of a tool for people to use for this purpose, and also the publication of your findings in a reputable genetics journal. Your hard work needs to be recognized by a wider audience. The same can be said for those working on haplgroups G and J etc. Progress is being made, but unless the information can be incorporated into Whits predictor or a separate tool, people are not going to know where to turn. You have genearated hundreds of posts on the subject, most of which I have saved, but the amount of if - then type of data is overwhelming and needs to be formalized and structured in a user friendly format. This could work out very well indeed since people could use your predictor to then order a specific SNP test so that when they enter the data in Ysearch there is a cross check and ultimately the element of certainty. The bottom line, we need to work together toward a common att!
ainable goal. There is enough expertise on this List to make virtually anything happen.

David F.



-------------- Original message --------------

> But their algorithm or their use of it is somewhat archaic (in this
> fast-moving field). What is their
> excuse of telling people who are obviously I1a that they are "I"? Certainly
> they can call a call of "I" a correct haplogroup assignment, but time is
> passing them by! They could also tell people they have the M89 mutation
> (indicating the all encompassing haplogroup F from the European
> perspective), but that's not very interesting as well.
>
> But maybe I will be pleasantly surprised, and someone will inform me that
> today FTDNA is routinely assigning people to I1a when they most very likely
> are?
>
> Ken

This thread: