Finding Your Biological Roots Stories from a genealogy detective:
uncovering family secrets, reclaiming the truth and uniting families

Wednesday, February 10, 2021

The Story of Bernadette - Part 15: My Last Hope for a Close DNA Match

puzzle pieces of family genetics
Six months passed with no new clues or close DNA matches for Bernadette. I was busy with three other cases during this time, managing to find two birth fathers, a half-sibling and a wayward grandfather. But Bernadette remained a mystery.

I dabbled with a new tool from Genetic Affairs, using Auto Cluster analysis to assemble matches into colored blocks according to shared DNA. A fascinating hi-tech method to quickly sort matches into groups, I found the Auto Cluster app much faster than using the manual LEEDS method. In order to access the Auto Cluster tool I first needed to register with Genetic Affairs then add my FamilyTreeDNA account and selected Bernadette’s profile.

The results soon arrived by email. I unzipped the file and watched as it magically pulled together tiny floating squares, each representing a cousin match from Bernadette’s raw file. Within a few seconds the screen resolved itself into an organized diagonal line of colorful blocks. I was left with 31 clusters and many floating gray squares on the outer fringes signifying DNA overlaps between two or more clusters. To put this into perspective, in a perfect situation you would have four major blocks of matches signifying a person’s four grandparent lines and then a smattering of smaller clusters. Of course you should never expect “perfect” in genetic genealogy. Therefore an average Auto Cluster usually contains 5 to 10 groupings - but 31 clusters! Obviously endogamy was again skewing the results and playing havoc with this new automatic sorting tool.

Auto Cluster was certainly a valuable and powerful genetic technology, but I needed more in-depth understanding before I could confidently utilize this tool in NPE research.

Bernadette’s case again returned to a state of inactivity, with the hope that something eventually would turn up and point me in the right direction.

By November of 2020, a year after Bernadette’s passing, inspiration struck in the form of an email from MyHeritage. Although I use MyHeritage only occasionally and was not a paying member of the site, I had taken advantage of their offer back in 2016 to upload my raw DNA file for free. Since they were new in the DNA testing market, MyHeritage was attempting to quickly build their database by accepting raw files tested with other companies. At the time, I thought “Great! Another DNA database where I can park my results and possibly connect with more cousins.”

On this particular morning, I received an email from MyHeritage with several distant cousin matches along with the country where they lived. I always found it interesting when a cousin from another country other than the United States appeared in my results. Usually I click through to view the match and attached comparison.

But today lightening struck - I had completely forgotten about MyHeritage for Bernadette’s research! How could I have overlooked this obvious database! I wasted no time, adding her raw DNA file that very afternoon.

Within twenty-four hours I received an email - Bernadette’s results were available and would I like to pay their $29 fee to unlock the account. I could barely contain myself. The MyHeritage database may very well be my last hope for a close match. Entering my credit card information, I clicked to charge the fee. Bernadette’s results appeared on my monitor.

I sat back in my chair with a gasp. With past cases, I have been happily surprised when a close match pops up and solves a birth mystery. But I wasn’t prepared for Bernadette’s results.

There it was! A close match sharing 25% of Bernadette’s DNA! This was the answer to her birth parents’ identity – and it was not what I expected.

The first name on the list I immediately recognized from the projected family tree I had built for Bernadette. It was the daughter of her older brother – Bernadette’s niece. Sharing 1,778 cMs of DNA, it was a 100% match! With this much shared DNA from the descendent of a sibling, there was only one conclusion – Bernadette shared the same birth parents as her older brother. Since her brother was born in Northern Italy not long after his parents’ marriage and he had immigrated to the United States as a young child with his mother, the brother’s birth parents were without a doubt, the same two people who had raised him. Therefore, there was no other conclusion: Bernadette’s true birth parents were indeed the parents who had raised her!

What a strange NPE case this had turned out to be! Bernadette believed for much of her adult life that she had different birth parents. And from conversations with her son, I understood this birth mystery belief was passed on to her spouse, children and grandchildren! So much so that Bernadette’s ex-husband had actually confronted a descendant of the supposed birth father, Frank, and demanded he acknowledge her as his half-sibling! Of course without any proof, this did not happen and the descendent dismissed the ex-spouse’s accusations as delusional.

But in the end, Bernadette was exactly who she was, the product of the two people who raised her as their middle child – there was nothing to uncover as there was no hidden birth secret!

It had been two years since Bernadette’s autosomal DNA was added to the AncestryDNA database. For two years I had tried to unlock a family mystery that was nothing more than a ghost! Although happy to conclude this case with an answer based on irrefutable DNA evidence, now I had to tell Bernadette’s children none of the lore surrounding their mother’s birth was real.

Although anticlimactic in its resolve, I felt Bernadette’s situation was a true learning experience of never expecting a pre-conceived outcome to a genetic genealogy case. Even though I suspected the DNA results were leading me in the direction of her real parents, Bernadette herself was convinced of something entirely different.

I took a day to compose myself after the final discovery before calling Bernadette’s son with the news. He was as shocked as I had been, having believed his maternal grandparents were someone else. Naturally he questioned the match with his first cousin (Bernadette’s niece), but soon resigned himself to the truth after I explained how I arrived at my conclusion through science-based facts confirmed by DNA. He had expected something different as had I.

In the end, the good news was that Bernadette’s children could now be certain of their maternal ancestry and of their medical history through their maternal grandparent lines. On further reflection, I felt it was probably best the truth had been found after Bernadette’s death. She may not have accepted this outcome, or perhaps Bernadette simply wanted to believe in an alternative reality.

As a search angel, I can only analyze DNA results and put together facts in the hope of solving family mysteries. I’ll have to leave speculation about Bernadette’s birth and life to her descendants. 

puzzle pieces of family genetics

Monday, February 1, 2021

The Story of Bernadette - Part 14: Comparing Admixture Results

DNA graphic
At this point, I decided to backtrack and revisit Bernadette’s admixture at AncestryDNA. The term “admixture” is another way of saying ethnicity results based on comparing a person’s DNA to specific gene pools to determine their ethnic mixture.

Predicting ethnicity percentages is an evolving “science” and somewhat speculative. Current testing can be certain of continents (such as a person’s ancestry being 75% European and 25% African), but narrowing the percentages to specific regions is still an approximation and should be viewed as such. Results vary between testing companies due to different test populations used by each company; different parts of a genome being tested; and differences between test companies in designating ethnic regions. Taking all of this into account, a person will never see the exact same ethnicity percentages when comparing DNA test results between companies. Your comparisons should be similar, but the percentages and regional designations will always vary.

You will also see similar but differing ethnicity results between siblings even though they share the same birth parents. The ethnicity results for my two siblings are close when compared to mine, but slightly different in percentages and minor population designations.

Although I was certain after examining Bernadette’s closest DNA matches and her ethnicity results that both of her birth parents were from the same endogamous community located in Northern Italy, I thought it was worth comparing her results against another DNA cousin with the same ethnic background. I knew their results would not be exactly the same, but a close approximation in their ethnicity percentages would provide further evidence of her maternal and paternal lines originating in the same close-knit community.

I happened to manage the Ancestry results for one of Bernadette’s DNA matches (a probable paternal 3rd cousin). Since I knew for certain that his paternal and maternal sides both emigrated from the same tiny village where Bernadette’s supposed birth parents came from, he was the perfect ethnicity comparison.

I accessed the cousin’s Ancestry DNA page and clicked on the pie chart under “DNA Story”. I copied his Ethnicity Estimate and pasted it into a Word document. Next I did the same for Bernadette’s Ethnicity Estimate, pasting it in the same document below her cousin’s. The results were so similar in populations and percentages that I thought I was comparing the results of siblings instead of cousins!

Plus Ancestry includes “Additional Communities” below their ethnicity results for populations with a high concentration of similar DNA to the tester. There were three communities listed for Bernadette, all accurate for this particular endogamous ancestral group. I scanned down to her cousin’s “Additional Communities” – and yes, they were exactly the same as Bernadette’s!

I felt this ethnicity comparison was substantial evidence that both Bernadette’s father and mother were native to the same small community in Northern Italy shared by her closest DNA cousins.

Earlier I proposed the scenario that Bernadette could have been the product of a younger uncle having a relationship with another woman from their endogamous community and the baby being given to the uncle’s brother and his wife to raise. But now I was beginning to suspect the parents who raised Bernadette actually were her birth parents. For an NPE case, this was a strange turn of events!

Still, I needed positive proof of her parents’ identity and the only way to confirm their relationship with Bernadette was to find a close DNA match. This mystery may prove to be unsolvable as Bernadette and all of her immediate family members had passed away. The only test subjects left that could establish positive proof were a few descendants of Bernadette’s brother and sister.

From previous analyzes of Bernadette’s closest AncestryDNA matches, I was confident her birth father was from the same family as the descendants of her siblings. This meant a DNA test from a paternal niece or nephew would without a doubt show up in Bernadette’s results.

I had to hope a descendant would eventually test with one of the companies where Bernadette’s results were registered. If a niece or nephew did appear in Bernadette’s results, their relationship estimate would be either a full niece/nephew or a half niece/nephew, depending on whether they shared one or both grandparents (one or both of Bernadette's parents). And if Bernadette’s father had actually been her younger uncle, the DNA match would indicate a slightly more distant relationship such as first cousin, once removed.

I had much less confidence a descendant would match her birth mother as I had found no DNA matches in any of the data bases with a close association to Bernadette’s maternal lineage. It was time to be patient and wait.

DNA graphic

Monday, January 25, 2021

The Story of Bernadette - Part 13: Let's Try FamilyTreeDNA

researching genetic genealogy online
By 2019 a number of NPEs had made contact asking for help and guidance. They reached out to me through various channels and I was happy to offer assistance as a search angel. Some of their cases were quickly solved through close family matches appearing in their DNA results. For others, I uncovered promising clues to their birth lineage. But concrete answers were still beyond our grasp, requiring further detective work on my part.

In comparison, Bernadette’s research was slow and dragging, yielding no useful information. I wasn’t giving up, but I must admit my enthusiasm was waning and I wondered if we would ever find conclusive DNA evidence for her birth parents.

With her health quickly failing, Bernadette was now in hospice. It had been many months since our last communication, although I was still in contact with her son.  Sadly in the autumn of 2019, Bernadette passed away at the age of 95. She had lived a long life, marked with many questions and suspicions about her birth family.

At this point, I could have stopped my investigation. But I felt Bernadette’s children and grandchildren disserved to know their maternal heritage along with possible medical history associated with their maternal grandparents. I pressed on, believing DNA would eventually yield answers to Bernadette’s family mystery.

The next tool in my genetic genealogy cache was FamilyTreeDNA. One of the oldest DNA testing companies, FamilyTreeDNA allows the upload of raw DNA files to their database. After your results are logged in and appear on your profile page, you then pay a small fee to unlock the rest of the company’s tools and perks. Since the fee is less than paying for another test kit, it’s worth the money to gain full access to their database along with other benefits.

Truthfully, I had little hope of finding a close DNA match here. FamilyTreeDNA’s database is small in comparison to other major testing companies, but it was worth a try. With Bernadette’s results included in a third database, we had exposure to a whole new group of potential DNA matches.

I uploaded Bernadette’s raw file and paid the fee to unlock her results. As I had suspected, there were no close matches, although the results were better than at GEDmatch. Her closest match was 168 cMs – a third cousin who I had already identified related to her paternal grandfather’s line. In total, Bernadette had 287 DNA matches at FamilyTreeDNA. Compare this amount with AncestryDNA’s database containing 3,069 shared DNA matches, with a few new matches being added every day.

One interesting plus included in FamilyTreeDNA results is X-match testing on the 23rd chromosome. AncestryDNA does not test for X-matches.

The X-match found on our 23rd chromosome is related to a person’s gender. Females inherit a pair of X chromosomes: one from their mother and one from their father. Males inherit an X chromosome from their mother and a Y chromosome from their father. This means there is the possibility of identifying the maternal lineage of a male DNA match through the X chromosome as he would certainly inherit this through his mother. However the male cousin’s mother would be inheriting two X-chromosomes one from each parental line, making it much more complicated to identify a specific lineage past this level.

Where an X-match can be most beneficial for an NPE is in the identification of a male sibling or half-sibling. If an unknown male sibling appears in FamilyTreeDNA results based on the amount of shared DNA, they will share an X-match if they have the same birth mother (or birth mother and birth father in the case of a full sibling) as the tester. If the match results for an unknown male sibling does not contain an X-match but still has the required amount of shared DNA to be classified as a half-sibling, then the tester and the male match share a birth father but not a birth mother.

Female DNA matches containing an X-match really aren’t helpful as the X-match appearing in results could have been identified with either a paternal or maternal line, opening too many lineages to trace. This was the case with the X-matches that turned up in Bernadette’s results – the matches were all female with the exception of one distant male DNA cousin.

Well it was worth a shot to look into this lone male cousin, although I knew he was too distant for the X-match to make a difference. He was not connected to a tree at FamilyTreeDNA. On a hunch I thought his results might also be at AncestryDNA. Although Ancestry does not show X-matches, perhaps this distant cousin could provide other clues. I switched over to AncestryDNA, logged into Bernadette’s results and did a search for the surname of her male X-match.

Success! There he was! And he had an extensive Ancestry family tree attached to his results along with a “common ancestor” match shared on Bernadette’s tree. From this information, I easily identified him as a third cousin, twice removed and a descendent of Bernadette’s paternal grandfather’s line.

Although this match was further confirmation of Bernadette’s paternal lineage, it offered no new information concerning his X-match. As I traced his tree through his mother’s line to the possible inheritance of the X chromosome, we again ran smack into the problem of endogamy! Going back just a few generations on the maternal side of this male cousin’s tree, I found surnames associated with all four of Bernadette’s supposed grandparent lines!

OK – so no close matches found at FamilyTreeDNA to positively identify birth parents. But evidence was growing that Bernadette’s birth father was most likely the man who had raised her.

Time to move on to the next DNA tool.

researching genetic genealogy online

Tuesday, January 19, 2021

The Story of Bernadette - Part 12: Uploading to Other DNA Databases: GEDmatch

DNA helix strands
Time was running out for Bernadette. She was now in hospice and no longer communicating with me. My next viable option in searching for her true birth parents was to upload her raw DNA file from AncestryDNA to other databases in the hopes of finding a closer match.

I contacted Bernadette’s son and received permission to download her raw file. Having access to her DNA file would provide more options in our search. Since a tester owns their own DNA results, testing companies make it easy for kit owners and managers to access and download DNA data to a ZIP file. This file can then be uploaded to other databases. In order to do genetic research for Bernadette, I had been assigned the role of manager for her kit, allowing me the ability to download her raw file.

Many NPEs are unaware that by testing with just one company, their matches are limited to only that particular company’s database. Unlike a decade ago, today there are many choices for autosomal DNA testing – and therefore many databases where a match can be found. If an NPE is serious about locating their birth family, they should make sure their results are in as many databases as possible, hence casting a wide net for DNA matches.  One way to do this is to download a raw file from your initial DNA test and upload it to other databases without having to retest.

But not all DNA testing companies allow you to upload a raw file from a different company to their database. Some will only accept testers who have purchased and completed a kit through their own company. Nevertheless, there are several options that welcome raw file uploads.

My first and most obvious choice was GEDmatch, an online service created to help others compare DNA results. Founded in 2010 long before the home DNA test craze took hold, GEDmatch is not a testing company. It is an online universal database where anyone can voluntarily submit DNA results from any testing company with the hope of finding genetic matches. Not as large as the AncestryDNA database, GEDmatch was originally used as the best way for NPEs and birth parents to find each other no matter where they were located or with whom they had tested.

It was a simple process to upload Bernadette’s raw file to GEDmatch. And in a short period of time, I was able to view her DNA matches. I kept my fingers crossed and hoped for a close connection.

Scanning the long list of DNA cousins I recognized a number of surnames associated with Bernadette’s paternal and maternal lines. This was not helpful as we know most of these surnames came from an endogamous community and they overlapped all four of Bernadette’s grandparent lines. 

Disappointingly, Bernadette’s closest match in the GEDmatch database shared only 43 cMs between three segments. This was a very distant cousin relationship, somewhere between a 4th and an 8th cousin - too distant to be of help. It was obvious there were no new clues to be had at GEDmatch to help with our DNA search.

Undaunted, I moved on to our next option.

DNA double helix

Photo Credit: Dmitry Sunagatov

Tuesday, January 12, 2021

The Story of Bernadette - Part 11: DNA Solves a Mysterious Baptismal Record

Researching Family Genealogy
To prove the birth father stated on Bernadette’s baptismal certificate was incorrect, I needed to disprove that Bernadette was related to the prominent businessman shown on her record. Let’s call him “Frank” (not his actual name).

As a side note: another interesting case had recently come my way involving the same “Frank”. How bizarre that this particular man seems to be connected with at least two questionable births within the same community. The story of how Frank plays a role in another NPE case will be told in a future blog post!

Returning to Bernadette’s DNA results, I decided to check if there was any connection with Frank’s descendants. Since I personally had a genetic connection to Frank (deceased many years ago); I knew the identity of his children and grandchildren; and I was aware that some descendants had tested with AncestryDNA, I used the search filter within Bernadette’s test results to locate the two descendants closest in age to Bernadette. If Frank had actually been her birth father, the two descendants should appear as Bernadette’s half-siblings.

Of course, we already know from our previous sorting of Bernadette’s DNA results, there were no close family matches. Any half-siblings appearing in Bernadette’s results would have immediately sent up a red flag and indicated a specific birth parent lineage.

None of Frank’s descendants were a close match to Bernadette. But were they cousin matches? The results of my search showed one descendant barely matched Bernadette with only 18 cMs, indicating a very distant cousin relationship. And the other descendant didn’t register at all as a DNA cousin, meaning that whatever DNA he may have shared with Bernadette was too low to show up as a match. (The minimum threshold for AncestryDNA is 8 cMs. Any match sharing less than 8 cMs could be a false or erroneous match and therefore it is eliminated from their results.)

The conclusion of these distant matches proved without a doubt that Bernadette was only very distantly related to Frank and he could not have been her birth father.

But while I had disproven Frank’s identity as Bernadette’s father, I still had no definitive proof that the man who raised her was her true birth father.

Without a close family match to use for comparison and both of her assumed parents deceased and unable to be tested, the results pointing toward Bernadette’s paternal side was still inconclusive and too distant. After much pondering, I decided there were two possible scenarios:

  •         Bernadette’s father was her true birth father
  •          Or a younger paternal uncle and brother to her father who lived nearby could have fathered a child (Bernadette) with a woman from their endogamous community. The illegitimate child was then given to his brother’s family to raise as their own. (This often occurred in families where an out-of-wedlock child was "adopted" by another family member and raised as part of their family.)

If the second scenario was true and Bernadette’s birth father was her uncle, she would still match the same cousins related through her paternal grandparent lines.

At this point in our search, we had made progress by:

  1.      eliminating Frank as the birth father
  2.       proving Bernadette’s baptismal record was incorrect
  3.       identifying both of her paternal grandparent lines as being correct
  4.       and zeroing in on two possible birth father scenarios.

Without any close, direct descendant matches to Bernadette’s paternal lines in her results, we had gone as far as we could with AncestryDNA.

Another disappointing discovery after sorting Bernadette’s closest ten Ancestry matches was apparently there were no direct descendants related of her maternal grandparent lines! Since the rest of her matches are at least 4th cousin or further in relationship, this obviously posed a problem in proving the identity of her birth mother.

It was time to try another tactic!

Researching Family Genealogy

Friday, January 8, 2021

The Story of Bernadette - Part 10: Using Ancestry ThruLines to Identify Paternal Lineage

AncestryDNA ThruLines Graphic
When it comes to genetic genealogy as with family genealogy, the best place to begin is always with what you know to be true. Since I had researched and compiled a three to four generation tree for Bernadette based on the parents who had raised her, this was my obvious starting point – the objective being to either prove or disprove their known identity.

I quickly put together a grandparent lineage chart using the surnames from Bernadette’s tree. And, of course, immediately a problem emerged. A number of surnames overlapped two or more columns on my chart! Since both of Bernadette’s assumed parents came from a historically endogamous community located in an alpine valley, it was obvious that all four of her assumed grandparent lines had intermarried for centuries and were interrelated. This is not unusual for an endogamous society, but it does pose a nightmare for a genetic genealogist attempting to sort out DNA results!

I scanned the first twenty to thirty matches and was relieved to see a few surnames that matched Bernadette’s paternal side, with several that coincided with her maternal lines. But remember, these were for the most part, all distant cousin matches sharing a small percentage of DNA. And it was apparent from the mutual surnames that Bernadette’s closest cousin matches also shared her endogamous heritage.

This was both a blessing and a curse! A blessing because Bernadette’s DNA matches were proving that one or both of her parents were from the same community, therefore narrowing our scope of possible birth candidates to her specific ethnic and societal group.  But also a curse because now we are attempting to sift through cousin matches most likely related to both Bernadette’s paternal and maternal ancestry with little distinction between grandparent lines.

Luckily Bernadette’s first few matches, although distant, were connected to family trees. Some even showed a “common ancestor” connection based on shared ancestors noted on Bernadette’s tree and their tree. Since I had linked Bernadette’s tree to Ancestry’s ThruLines tool, it only took a few minutes to identify the common relative of her closest matches along with their relationship to Bernadette.

Of Bernadette’s top ten matches, four were direct descendants of her assumed paternal grandmother’s line, three were related to her assumed paternal grandfather’s lineage, and three were related through both of her assumed paternal grandparents. Since Bernadette’s closest match (sharing 155 cM) was connected to an extensive tree, it was easy to determine their genealogy relationship as second cousin, once removed with Bernadette being the elder of the two. 

DNA was certainly pointing us in the direction of Bernadette’s birth father being the man who had raised her, which would disprove her questionable baptismal record.

AncestryDNA ThruLines Graphic

Monday, January 4, 2021

The Story of Bernadette - Part 9: Using the Shared Matches Tool

Graphic of DNA double helix
Now let’s use the “Shared Matches” function in Ancestry to identify other cousin matches related to an exact grandparent line. Click on a known match that has been assigned to a color coded group. Next click on the link “Shared Matches” found at the top of the page. This will bring up a list of cousins you have in common with this match and with whom you share a portion of DNA. If you’re not using AncestryDNA, you can still make use of a shared match function, as all testing companies offer this option. Matter of fact, I consider this filtering tool to be the most important component of any DNA testing result!

If the match is a close family member or a first cousin, you may share hundreds of matches through various family lines. This initial sort will allow you to designate matches as either maternal or paternal – but it may not offer enough information yet to sort into a specific grandparent line. This is only possible if you can identify a specific surname or a known family relationship that can with certainty be associated with a particular grandparent lineage.

But what happens when you either don’t know or aren’t sure of either paternal or maternal lineage? This was Bernadette’s predicament – all of her grandparent lines were in question.

DNA double helix graphic