Same DNA file, two answers: the DNA industry won't explain the gap

Disclosure: True News is a Bagley Enterprises platform, and this story is about the company's owner, Len Bagley, and his independent research. It was reviewed under the same standards as every story we publish. Bagley's findings are presented as his own and attributed to him, and the companies' responses are quoted from their written replies to him.

Many Black Americans who take a consumer DNA test receive the same kind of result: 80 percent or more African, often broken down by country, and sometimes by named ethnic groups such as Yoruba, Mende or Mandinka.

Len Bagley's results carry that label. The bulk of his family matches sit in the American South and Mid-Atlantic: Virginia, the Carolinas, Georgia and Florida. The rest are in Europe, including Spain, Germany and Ireland, and in Canada and South America. He has none in Africa, and none in the Caribbean.

"They tell us we are 80-plus percent African. Some even break it down to tribes," Bagley told True News. "But we have no family matches on the continent or in the diaspora. How do I come from these people but am related to none of them?"

Bagley, a Philadelphia business owner and independent researcher, has put that question to the DNA testing industry in writing since November 12, 2025. This follow-up to True News' story on his research sets out what he asked and what the companies said.

One file, two outputs

Bagley's AncestryDNA results show two outputs from the same DNA file, and they disagree. The ethnicity estimate says 85 percent West and Central African. The company's own match-based "communities," which group testers by the relatives they share, place every one of his family connections in the American South and Mid-Atlantic: Virginia and Southern Ohio African Americans, Early North Carolina African Americans, Early South Carolina African Americans, Georgia Coastal Plain to Central Florida African Americans, Deep South African Americans, Southeastern United States African Americans, and Piedmont Plateau African Americans. None point to Africa.

Living DNA reported about 87 percent West and Central African ancestry, naming Yoruba, Tikar, Mandinka, Mende and Semi-Bantu reference groups, plus 13 percent England and Wales. African Ancestry assigned his maternal line to a modern West-Central African group. Its paternal-line results letter, dated July 14, 2022, said that after comparing his DNA to its African lineage database and finding no match, it found "identical, 100% matches with the Y-chromosome DNA of a European man," with paternal ancestry among people in Spain and Portugal: "There are no African men that match your DNA sequence." The letter notes that the result "is for only one branch of your family tree," and Bagley notes that it describes one paternal line out of many. He also stresses that "European" in such results describes where a genetic lineage is found today, not skin color: DNA from Europe's early hunter-gatherers, such as Britain's 10,000-year-old Cheddar Man, indicates they had dark skin.

Bagley's finding is that living relatives are the part of a DNA test that can be checked against birth certificates and family trees, and his trace to Virginia, the Carolinas, Georgia and Florida. The percentages are model estimates built on reference panels of living people. In Bagley's analysis, none of the companies has shown how its panel handles a population with no reference group of its own.

His claim is specific. The label "African, full stop" has been applied to people whose own matches point nowhere but here. No company has checked it against the right comparison: ancient DNA from the people who lived on this land before them.

If they can name the tribe, they can find the cousin

The companies sell this result at a high level of detail. Living DNA names five African reference groups. AncestryDNA divides his African ancestry by region and country. Bagley says that detail cuts against the companies' own excuse.

A company that claims to tell a Mende ancestor from a Yoruba ancestor in his DNA is claiming fine-grained resolution. That resolution, he says, should produce living relatives. His ancestry is also not supposed to rest on one line. By the companies' own accounting it is spread across many groups, and each group is another chance to match someone alive today. A person who is more than 80 percent African, divided among that many groups, has far more opportunities for a match on the continent than a person descended from one place. He has none.

The record, date by date

Bagley says the companies' answers are "telling," and describes them as a shield rather than an answer. Below is every exchange, with each company's words exactly as written to him.

November 12, 2025: In a live chat, Bagley asked AncestryDNA to escalate his questions to its population genetics team. By his written account, the support desk replied: "The White Papers contain all the information which we make available to the public. A request for further information or statement will not be possible, I'm afraid."

November 12, 2025: The same day, Bagley sent AncestryDNA, Living DNA, African Ancestry and GEDmatch the same evidence packet, which he titled a Population Reference Bias Audit. He asked each company whether its reference panels include ancient American genomes, how its software classifies DNA that matches no reference group well, and whether it would review his results against ancient samples.

November 12, 2025: Living DNA replied: "I have forwarded this onto the science team to take a look into. Once i receive i response, i will be in touch." African Ancestry sent an automatic reply, then asked for his seven-digit kit number. GEDmatch and Verogen did not reply.

November 13, 2025: AncestryDNA's Executive Office replied: "our research and scientific development divisions are focused on matters specific to their division. They do not handle direct customer correspondence, nor can this office provide technical clarification beyond the information already published." It added: "Because AncestryDNA focuses on information relevant to genealogical research, our panels are based on modern indigenous genetic segments whereas other companies may focus on ancient population origins."

November 13, 2025: Bagley wrote back asking for confirmation that his packet had reached the science team, a written statement on whether this is a known limitation of the company's model, and its policy for correcting such limitations.

November 17, 2025: AncestryDNA's Executive Office quoted its terms and conditions, which describe results as being for "informational, recreational, educational, and personal research use," and said: "Ancestry currently has the largest consumer DNA database with 25+ million samples and use a reference panel with over 185,063 samples from across the world to determine ancestral origins. Additional information beyond what has been publicly released is considered proprietary and is not released upon request."

November 20, 2025: Bagley asked Living DNA for an update. Living DNA replied: "Your review has been ticketed for the science team look into. As soon as i receive a response, i will sure to be in touch." The same day, African Ancestry sent a reply that contained a greeting and a signature but no message.

April 20, 2026: After five months without an answer, Bagley wrote to Living DNA again. That day, the company's partnership manager relayed a response from its science team: "We are confident in our result, and it is not a marginal or ambiguous one." On ancient-DNA tools such as GEDmatch's, it said they "are community-assembled resources and are not peer-reviewed instruments," and that "All living humans carry alleles that show measurable similarity to ancient genomes." It called reference compression "a genuine concept in population genetics" that applies to under-represented groups, "like some Native American and Oceanian populations," and said: "This is the inverse of your situation."

April 20, 2026: Bagley asked Living DNA whether African Americans were used to build its African reference groups, and wrote to Verogen, the company that owns GEDmatch, asking why six ancient genomes on its site had been set to private. The letter to Verogen's legal department was returned as undeliverable, so it was never received.

April 21, 2026: Living DNA answered: "African American individuals are not included in West or Central African reference populations. Inclusion in any reference panel requires documented multi-generational ancestry in a specific region." On the missing relatives, it said "the absence of detectable cousin matches on the African continent is expected," because "African populations are substantially underrepresented in consumer DNA databases globally," and "the geographic distribution of your matches reflects the geographic distribution of who has tested, not the geographic origin of your ancestry." It said its "position on your results remains unchanged" and closed the correspondence.

May 3, 2026: Writing for PAAN Ministries, an organization he helps lead, Bagley sent AncestryDNA and Living DNA formal notices demanding answers within 14 days. Living DNA sent an automatic acknowledgment.

May 6, 2026: AncestryDNA's Executive Office replied: "Ancestry does not provide sworn statements or legal declarations regarding DNA Ancestral Regions." It quoted its terms, which provide its services "on an 'AS-IS' basis, meaning without any guarantee or warranty," and said: "GEDmatch allows users to compare their DNA to ancient archaeological samples which often show high 'noise' or coincidental matches due to shared deep human ancestry."

As of this story, African Ancestry has not addressed his questions, and GEDmatch and Verogen have not responded.

What the record shows

Bagley points to three things the companies' own words establish:

• The panel is secret. AncestryDNA calls everything beyond its published figures "proprietary" and "not released upon request." Customers are asked to accept a percentage without seeing how it was built.

• The companies' own products disagree. AncestryDNA's ethnicity estimate says Africa. Its match-based communities say the American South. The company has not explained the difference.

• Black Americans are not in the African reference groups. Living DNA confirmed that "African American individuals are not included in West or Central African reference populations." Bagley says that means the African categories were built from other people. A result of "African" for a Black American is then a statement about resemblance to those people, not a measurement of lineage.

The companies dismiss ancient-DNA tools as "noise" and "not peer-reviewed instruments." Neither has said it ran the comparison Bagley asked for, and AncestryDNA's terms disclaim any warranty on the result.

The missing relatives

Living DNA says "the absence of detectable cousin matches on the African continent is expected." Geneticists add that the chance of detecting a distant cousin falls with each generation. 23andMe estimates about 45 percent for a fourth cousin, 15 percent for a fifth and under 5 percent for a sixth or more distant cousin.

Bagley's answer is that his own match list shows distance does not explain it. Living relatives in Spain, Germany, Ireland and elsewhere in Europe sit at the same distant, hard-to-detect depth that supposedly erases his African ones, and the same software finds them. It also finds relatives in Canada and South America. It finds none in Africa, and none in the Caribbean.

He also argues that "who has tested" does not hold up on the numbers. Roughly 2.5 million African-born people now live in the United States, most of whom arrived after 1990. They live in the same cities where testing is common, and they do find relatives on the continent. Each person also has thousands of distant cousins, about 4,700 fifth cousins on one simple model. Even at a low detection rate for any single cousin, he says, a population descended largely from West and Central Africa should turn up some relatives there. His do not.

When the percentage and the matches disagree, Bagley says, the matches are the part that can be checked against records.

He adds that the same kind of DNA segment matching has been used to detect a single contact between Polynesians and South Americans about 800 years ago. In his view the method can answer his question too, and the only missing step is that no one has run the comparison.

What he is demanding

Bagley is asking the industry for three things:

• Publish how each reference group was built and who was excluded.

• Let the software say "undetermined" when a person's DNA fits no group well, instead of assigning every result to the nearest category.

• Run a transparent comparison of Black Americans with documented roots before 1900 against ancient DNA from the Eastern Woodlands and the American Southeast, with the method published.

Until that comparison is run in the open, Bagley says, the percentage on a test result cannot be treated as the final word on where a family comes from.

In part three, True News examines the ancient-DNA tests Bagley ran on his own file, and what they show about the resolution behind the companies' claims.

On October 7, 2026, True News asked AncestryDNA, Living DNA, African Ancestry and GEDmatch to comment on this story by October 14. Any responses will be added here.

Bagley's research is available through his Academia profile, linked with this story.

Originally published by True News: https://truenews.bagleyenterprises.net/stories/15