Re: Ancestry DNA – Why Low Confidence Matches Matter
Interesting article:
http://dna-footprints.com/482/ancesty-dna-confidence-matches...

Very interesting article! I don't completely follow the technical terms, but it clearly highlights the difference in the ways the three major companies are analyzing the data and choosing to select "matches". It not only helps explain why the very low matches on ancestry.com are important and seem to produce a lot more vertified results than one would expect, but also helps explain why people who test at all three companies might not get matches to the same people testing at all three.

Did this thread just get moved? I thought I had posted on the more appropriate board ("DNA Autosomal") instead of the less appropriate one ("DNA General Research"). Oh well. Sorry.
I don't follow it completely either. As others have, they mention that Ancestry uses "phasing": "Unlike other companies, they phase the data from the get go — meaning that smaller segments are more confident and that they have many fewer IBS segments." This must be a different kind of phasing from the kind we've discussed earlier on these boards. ??? In any case, you're right. Until now, it has not made sense to me that we seem to get as many paper-supported matches from the low-confidence end as from the high-confidence end.

Very interesting article. Thanks! Yes, I agree that because your link concerns autosomal dna only, it does belong in the autosomal sub-forum, but who knows why things get moved around here?

Sherlburl, I wonder if this explanation has to do with the actual placement of the matches in the relationship categories. In other words, this Mb measurement is what places a match higher or lower in the groupings. For example, I had 2 third cousin predictions, and when I received a new one, it was placed in the middle of the two. At Ftdna, we can filter our matches by relationship, shared cM, or longest segment. From your blog link, I'm assuming that this placement is Ancestry's filter of high to low. no?

PS I still haven't been able to find our match. In light of your link, this may be something I need to look at. :)

I can't find our match either, Marthe. :) I think you're right about the rankings. I have 4th cousin matches to a man and his mother. When her results came in, she appeared two matches above him. I have another parent-child pair of matches and they are consecutive (no matches in between).

sherburl -

Yes, you originally posted on the autosomal board, but didn't mention autosomal in either your subject line or in the body of your post, so the post was moved to General. You then edited your post to include the word autosomal.

This passage relates a concern that has been previously raised. We will need some type of conversion utility to use the Ancestry data on gedmatch or other sites:

"AncestryDNA has said they will release the segment information sometime this year. When they do, the information will still not be directly convertible to compare to other companies. You will need to convert the Mbp to cM to be meaningful. But you should be able to do a direct comparison to your AF location."

Jim

Jim,
Would you please explain more? What is an "AF location"? Thanks for your help,
Lou

What it means is that you should be able to identify and compare the segment locations with your matches on Ancestry, but because they are using a different type of measurement, you will not be able to compare the specific segments from your Ancestry matches with those from other companies.

This will be a big limitation unless they provide some kind of conversion program.

Jim

Raw data files from 23andMe and FTDNA give the SNP location in Mb units, so that shouldn't be a problem. The companies and utilities like GEDMatch have routines to report the distance between two points in cM.