I (female) had an unexpected half sister match (1,866 cM across 49 segments) in Ancestry.com. We are about the same age so I'm assuming sibling. Upon closer examination I realized my brother(A) was actually a half sibling match (1,696 cM across 72 segments) and not a full brother. This explains why I didn't match anyone related to our father unless it was a descendant of our mother too. And there are dozens of matches to my new half sibling's family to me. My apparent biological father lived two miles from my family. I'm okay with all of this and it actually explains a lot of my childhood but my brother is livid. He thinks it's all a scam or an error. Also, the daughter (my niece) of another brother (B) matches the new biological family which would point to this brother being a full siblings having the same (new) biological father. This brother's DNA is not in the system. He won't add it because he too thinks this whole scenario is crazy and it is maligning our mother! Both fathers are deceased, hence no DNA from them.

This would have to be a lot of errors! I'm new at this and don't know how to go about verifying these results. How can I prove what's in Ancestry.com is legitimate? Has anyone ever had a "wrong" match before? And wrong when there are this many matches?

Too make matters worse, my brother A and I sent our DNA to another company of his choice (Biogene) and it came back as 97% full siblings which indicates Ancestry.com is wrong. I don't know how this could be. I can't find out much on Biogene so don't know if it is reputable or not.

My new half sister hasn't responded so I'm assuming she is trying to reconcile the connection. It is not my intention to disrupt anyone's life so I don't think she's going to be helpful.

Wow, so much to say! First, someone tell me if I'm posting at the right place. Do I edit my original post in this manner or add a comment?

We did cheek swabs with Biogene. My brother ordered the tests. He was looking specifically for a sibling test. He doesn't believe cM's will tell him anything. We don't live close to each other (thank goodness, excuse the editorial). I've tried to call the company for an explanation of the test and why I don't see cM's on the result. They will not talk to me. I've tried to do research on Biogene but can find very little that's not related to their own PR. My brother thinks they are the number one DNA company out there.

My brother will not use the companies you suggest as he says it's all a con game and they just want your money. I've pleaded with him to use GEDmatch to no avail.

I have called Ancestry. They were very cordial but really not much help. Just gave me links to their online info.

As far as my new match possibly not being a half sister, I've examined this inside out. I'm definitely new to this game but I'm trying to be organized and methodically and definitely leave the emotion out of it (except for the stress my brother is causing!).

I am not related to any of my father's (my surname) relatives who are in the system. We were able to get my mother's DNA before this scenario popped up. She is not related to any of my new unexpected matches. My niece from brother B matches the same people I do (with less cM's) and also does not match my father's relatives. This is less officially but a pic of possible biological father as a young man looks amazingly like brother B.

Can I add a screen print of the results from Biogene here if I omit the header with names?

Thanks to everyone who has taken an interest in helping me. I don't have a clue how to rectify this.

My niece who is the daughter of brother B whom I believe must also have the same suspected biological father as me has this relationship to me: 1,652 cM across 62 segments

I can't get the prt scr to work here for the Biogene test results, Any suggestions?

Biogene report enter image description here

I can't get the cM matches between my niece and my potential half sister. No one is speaking to me!

  • 2
    The differing results are very unusual. Maybe let the companies know about the contradictory results? If there was no error, one of you might be a human chimera. Please let us know what you eventually discover. Commented Dec 22, 2018 at 4:11
  • 1
    @EllenSpertus I think your comment contains just enough content to be worth posting as a minimal length answer.
    – PolyGeo
    Commented Dec 22, 2018 at 7:46
  • I am a novice to this website as well. Don't understand the "posting as a minimal length answer". Should I have posted elsewhere? Keep me straight!! Thanks
    – J Joan
    Commented Dec 22, 2018 at 15:35
  • 1
    @JJoan, what you've done is good, and "PolyGeo" wasn't asking you to do anything; they were addressing "EllenSpertus" in reference to her first comment.
    – RobertShaw
    Commented Dec 22, 2018 at 18:16
  • @JJoan, I would like to ask about your Biogene testing. Did you two send biological samples, such as cheek swabs? What type of test did you two order? What was the wording of the result?
    – RobertShaw
    Commented Dec 22, 2018 at 18:36

6 Answers 6


Errors of this magnitude are possible but very very rare.

I'm sorry I don't have the documentation but there was discussion on at least one of the lists/groups for genetic genealogy about a case where an AncestryDNA customer was told that someone matched her as a parent. Huge surprise. Turned out to be a bizarre error and the two people weren't even related. It was well documented and verified by some of the top people in the field who participate in these groups, and it was not the only case.

I don't know of any cases where two people are in fact related but where AncestryDNA put the wrong amount of cM in the match. There are plenty of times (for all the companies) where the suggested relationship is wrong (because there's usually more than one possibility) but, in your case, neither of these numbers is even close to a full sibling (which should be at least 2200 cM).

I'm not familiar with Biogene so I can't comment on them, but it worries me that you got a result saying "97% full sibling" which is not an actual number of anything but a confidence prediction. You need a company that gives you cM numbers and segment information.

Where to go next.

I have two suggestions. I hope you will do both:

  1. Take your raw data from Ancestry's and your brother's and put them both on Gedmatch. It's free and will only take a few days to be ready to use. Do one-on-one matching to look at the segment data. Full siblings will have multiple segments where they match on both sides. By that I mean, segments that are fully identical because they share one stretch from mom and a different but overlapping one from dad. Half siblings will only have half identical segments.

  2. Test with a new well known and trusted company doing genetic genealogy. I suggest 23andme or FTDNA. There are Christmas sales going on! One advantage to 23andme is that you'll get a mtDNA halpogroup as part of the test and this can be helpful if you're not 100% certain you share the same mother. What would be even more useful for you though is the Y-DNA halpogroup. Your brother will get a simple one for free at 23andme but I'm guessing from your username that you're female. If you have a Y-chromosome, then I strongly urge you to use 23andme as the quick and dirty check and then followup at FTDNA (with a separate Y-DNA test) if it matches. Or you can just go for FTDNA to start.

More than a sibling.

Your new match might not be a half-sibling, even if all the tests are correct. 1,866 cM is quite consistent with an aunt relationship too. While the fact that you're the same age does point more to sibling, aunt/uncles are sometimes the same age as their niece/nephews. Even in families where everyone is a full sibling. All you need is a family with a 20 year spread between the oldest and youngest sibling and that oldest sibling has kids fairly young. If there's a teen pregnancy somewhere, it's even more common.


Adding another answer in order to address two new additions to the original question, each of which could be questions on their own.

1) The Biogene test is not the same test as we use for genetic genealogy. It's a paternity test, so to speak. It looks at a few important markers and uses them to determine very close relationships.

Bio-Gene DNA Testing’s Legal Relationship Test and Home Paternity Test both examine 24 Genetic Markers in a person’s DNA, checking for genetic markers between the suspected father and the child or children.

And wow this company gets awful reviews. For customer service and unclear results. I'm not finding anything saying they're inaccurate but it's not certain if they do a good job differentiating between full and half siblings. Their main focus is differentiating between children/parents and unrelated people.

I trust AncestryDNA's cM count more than I trust this place, but all I can say is they contradict. There is no way 1,696 cM is a full sibling. So unless you made a mistake with the number, Brother A isn't your full brother.

2) Your niece is your full niece.

My niece who is the daughter of brother B whom I believe must also have the same suspected biological father as me has this relationship to me: 1,652 cM across 62 segments

1,652 cM you'll notice is almost the same level of match you have with Brother A. Your match with Brother B is going to be much bigger than that, right?

Looking the Shared cM Project chart:
Full niece: 1349-2175 cM range with the mean at 1750 cM.
Half niece: 500-1446 cM range with the mean at 891 cM.

I'll note too that, just because we know there's some weird stuff going on in your family tree, that your niece is also in the range to be your half-sibling. Her father, Brother B, refuses to take a DNA test. Now, lots of people refuse and there's no secret behind it. They just don't wanna do it. But it makes me wonder.

What is the match between your niece and the woman you don't yet have a relationship for? That may help narrow this down.


DNA is not something one can do easy-peasy relationship labels with, especially when it comes to conflicting stories. The Match List will give you the 'norm' label for that DNA range. And it is a range! Each relationship has a minimum and a maximum DNA count for CM that can fit for that relationship. However, if you add in endogamy (i.e. cousins marrying), twins, 2 brothers in one family marrying 2 sisters in another -- the range will change. The other things to consider are adoption, sperm donors, out-of-marriage children. So you see it is a subject to approach carefully.

The idea is to find out the most you can on your end first, so you can make sure all your ducks are in a row with no doubts. Sometimes you have to wait a long time for more people to post DNA before you can come to a VALID decision. You have to put your scientific brain on to puzzle this out because though DNA doesn't lie, it can go through some very strange round-about routes. The Articles below are by the best. But don't just look at the one person you are concerned about...look at all the OTHER connecting people to that person too as their DNA will play a big part in figuring out the puzzle.

https://dna-explained.com/2019/04/03/full-or-half-siblings/ https://thednageek.com/science-the-heck-out-of-your-dna-part-2/ https://dna-explained.com/2017/12/27/concepts-sibling-and-twin-dna-matching/ https://dna-explained.com/2018/08/21/proving-or-disproving-a-half-sibling-relationship-using-dnapainter/ https://dna-explained.com/2019/04/04/things-that-need-to-be-said-adoption-adultery-coercion-rape-and-dna/

The second chart in this article will help. https://isogg.org/wiki/Autosomal_DNA_statistics

Ok, that should definitely get you lots to explore and put you on a better path to success in your quest.


Systematic errors are extremely unlikely. Almost the only possible error scenario is if the test actually processed the DNA of a different person. This would most likely be if you (probably not Ancestry) mixed up the kits when multiple people took the test at the same time.

If either you or your DNA match has ever been a bone marrow donor or recipient, you can get unexpected results. In at least one publicized case, a DNA test on a bone marrow recipient yielded results and matches from the bone marrow DONOR.


I agree with the long answer above by user10321. I however would like to add that through the use of 23&me by family members I place little faith in that company. Both members had used Ancestry DNA testing in the past and through my extensive 10 year journey through creating a family tree on Ancestry followed by DNA testing of both my husband and I a clear and validated ancestral history has been created. Substantiated by many proofing methods. The 2 people who used 23&me as a follow up were my husbands close relatives who believed there was a strong presence of French ancestors. The report from 23&me stated they were descended from Marie Antoinette. I decided to research her further and discovered... that was not her real name, she was not French, she had 4 children with the mad king of France of which only 1 lived past the age of 10, and this daughter had a child that died very young. So Marie Antoinette had no grandchildren that lived long enough to have descendants. Genealogy is arduous and complicated study but it offers an interesting study of one's history and story. However, having helped 4 adoptees find their roots I know from this experience not everyone wants to know this story and many are fearful of others knowing their ancestors.


Hi I believe ancestry can make mistakes. I am adopted and have conflicting dna from ancestry as well as dna tribes which seems more detailed by th e way... one way or another there is a flaw in the system bc one says I am iraqi whereas th e other comes up with Fijian Indian as my closest relatives...go figure that one!!..

  • 2
    Hello, welcome. It appears you are referring to DNA ethnicity estimates, not DNA matches which this question is about. It is not uncommon for different companies to have somewhat different ethnicity results, based on their DNA databases and reference samples. As more people test, the estimates become more accurate. Adding sources to your answer will give it more credit - you can see the upvoted answers on this question have references, while the downvoted ones do not. Feel free to edit your answer to clarify.
    – Harry V.
    Commented Nov 8, 2019 at 3:44

Not the answer you're looking for? Browse other questions tagged or ask your own question.