23andme DNA result help needed

+4 votes

23andme Matches and Reports trying to understand them...  

I am working with someone who tested with several labs including 23andme.  So these questions relate at least to the tests from the same lab.   

On their reports,  23andme has a designation of half on some chromosomes and complete on others.  (Complete not the word full)

In various places on their website I found the following information:

>23AndMe’s criteria for matching segments are a distance of 7cMs and at least 700 SNPs for the first segment and for segments of other people you are sharing with. 

>The start and stop positions of the 23AndMe test are usually rounded to the nearest millionth base pair in Family Inheritance, while the length of matching segments is rounded to the nearest tenth of a cM.  

>Relationship Average % DNA Shared

Full Sibling ...............  50%

Half Sibling...............   25% 

>When you are dealing with siblings, some of the values are fully identical regions and some are half identical regions. The fully identical regions should be counted twice, but in some cases, they are not, they are only counted “as” at match, not as a match to each parents, so counted twice.   (my note:  is complete fully identical regions?)

>In some cases, where you match both sides is counted at 1 match and in some cases, it’s counted as 2, one for each side. The easy way to tell if someone is a full or half sibling is that full siblings have fully identical regions (FIR) and half siblings do not, because they only match ONE parent.  (the report does not indicate FIR but has half and complete under a column heading of "Identity" once again complete not full)

In our situation:

>"B" and "C" share 52% DNA in common; 

>"B" and "A" share 47.7% in common; and 

>"C" and "A" share 54.4% in common. 

By total CMS they should be full brothers: 

>"A" to "C":  2871 cms; 

>"A" to "B":  2643;

> "C" to "B":  2897.  

When these 3 people are matched to a "cousin D" they match as:

>A to D      354 cms   12 segments  using Bettinger tool  Half 2ndCousin Half 1st Cousin, 2nd Cousin, Half First Cousin 1 Removed, 1st Cousin 1 Removed, Half First Cousin 2 Removed, 1st Cousin 2 Removed, 

>B to D     917 cms  19 segments using Bettinger tool  1st cousin

>C to D     713 cms   15 segments using Bettinger tool  1st Cousin or Half 1st Cousin, 1st Cousin 1 Removed  

All of which is confusing because there seems to be a more distant relationship with one of the siblings but if they are full siblings shouldn't the relationship with each one be basically the same?

(Jillian is that easier to read?)

in Genealogy Help by Laura Bozzay G2G6 Pilot (854k points)
edited by Laura Bozzay

Laura, your message is very difficult to read, mostly due to formatting issues created by cut-and-paste content. Would it be possible to edit it so it's easier to see what it was intended to look like?

One comment on substance. You say "23andme has a designation of half on some chromosomes and complete on others." That's something that appears when sibling data is compared. If the two siblings got the same chromosome segment from one parent, it's a half-IBD match. If their other copy of the same segment (for example, the first half of chromosome 1) from the other parent is also the same, it's a full-IBD match.

I see that the three siblings match on roughly 50% of their DNA. Do all three siblings A, B, and C share some "full-IBD" matches? (These indicate full siblings.)

It's normal for different cousin pairs to have different amounts of matching. But the match between A and D is unusually weak for full first cousins, which would make sense if A is only a half-sibling to B and C.
The report is saying "half" and "complete"  it does not say "full".  So I am wondering if complete under the title of identity on their report is what they describe as full in their help text....  inconsistency of language between reports and help text docs is not helpful...   I do not want to assume this is the same meaning and because the relationship with a "cousin" to these 3 supposedly full brothers  does not show that way, it is confusing and we are trying to figure it out.
Thanks for reformatting the question, Laura. That makes it much easier to follow!

1 Answer

+3 votes
Best answer
I think that you and your correspondent may be working from the DNA Relatives page and comparing people's DNA one pair at a time. When you do that, there's a graphic at the bottom of the page under the heading "Shared DNA." The graphic shows the 23 chromosomes (including X) and color codes them to show segments that match and segments that don't match. Your description of "half" matches and "complete" matches appears to come from the legend that explains the color coding on that graph; the categories are "Half identical," "Completely identical," "Not identical," and "Not enough information." (Apparently they've changed their wording. I guess they decided that the old "Half IBD" terminology was misleading.)

You'll get much more powerful information about the matches -- and you are likely to get answers to your questions -- if you go back to the top of the DNA Relatives page and select "DNA Comparison." On the DNA Comparison page, you can do more detailed comparisons of one member's DNA with up to 5 other members' DNA, and see a graphical display of the matches, as well as numerical data on each match. (That graph also codes the matches as "Half identical" and "Completely identical.")

Try selecting person A as the first person to compare, then enter the names of persons B, C, and D in the next three boxes, and compare. If A, B, and C are full siblings, the color coding for persons B and C should include both Half identical and Completely identical matches, and person D should have only Half identical matches. If A has only half-identical matches with B and C (this is unlikely), then A is a half-sibling. Save a copy of the results to consult when reviewing the next set of comparisons.

Now select person D as the first person and compare D's data to A, B, C. Line the results up against the results from the previous comparisons. With the data from these two comparisons, you can catalog the places where the matches are. If A is a full sibling of B and C, but matches D on only 354 cM, you might discover that a large fraction of D's matches to B and C are on chromosome segments where A has no match to B and C.
by Ellen Smith G2G Astronaut (1.6m points)
selected by Laura Bozzay
Thanks I will ask her to do this.  I do not have access to 23andMe and am not as versed in their reports as I am on FTDNA and GedMatch reports which make sense to me.   This just seemed odd with the big difference in CMS among the 4 of them.
Yes, that 354 cM match is astonishingly low for first cousins.

The DNA Comparison feature on 23andMe is similar in many respects to comparing DNA on GEDmatch. But I think some 23andMe users don't manage to find it.

And I note that the term "full IBD" that still appears on some of the 23andMe help pages means the same thing as "Completely Identical."
Great!  I suspected that but did not want to assume since it would have led to different conclusions if I was wrong...  

I asked the "cousin" to do what you suggested and see if it clears up the confusion.  I appreciate your insights.  I also just heard they are going to upload their tests into GedMatch which means we can see how they stack up there.

Related questions

+6 votes
3 answers
+13 votes
1 answer
+4 votes
3 answers
+6 votes
2 answers
+5 votes
2 answers
232 views asked Mar 9, 2023 in The Tree House by Richard J G2G6 (10.0k points)
+15 votes
10 answers
+6 votes
0 answers
+13 votes
3 answers
+7 votes
0 answers

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright