Wednesday, October 15, 2025

Can ChatGPT Analyze Your Raw Data Zip file? Yes it can!

 When I was analyzing my new AncestryDNA ethnicity results I wondered if no calls could have affected my ethnicity results? Could the missing ethnicity be due to no calls? I asked ChatGPT that question. I was surprised when I got the answer that I could upload my raw data and check for no calls. I found out I only have 1.64% no calls. According to Chatgpt this is a low percentage of no calls and they are spread throughout my genome. This shouldn't affect my ethnicity results or matches. 

No calls look like this in your DNA raw data:

Correction:

I had to make some corrections to this post. I had actually used a 23andMe raw data file by mistake when I first had ChatGPT search for no calls. I now had ChatGPT search an AncestryDNA file for no calls. I have more no calls at AncestryDNA. According to ChatGPT it has 1.81% no calls. More than 23andMe. I had to tell ChatGPT to look for 00 separated, otherwise it can't find the no calls. Hopefully it did process the file correctly. Eventually I'll figure the percentage myself to verify. Below is what ChatGPT told me about these these no calls and their impact:

🧬 1. DNA Matches (Cousin Matching)

  • Small impact overall: The algorithms (at Ancestry, 23andMe, MyHeritage, etc.) look for long stretches of identical DNA (identical-by-descent segments). A missing SNP here and there usually doesn’t break those segments because they use overlapping windows of markers.

  • But: If no calls cluster in certain regions, they can slightly shorten detected segments or even cause very small segments to go undetected — potentially reducing the total shared DNA a little.

  • In practice, you might lose a few distant matches (like 6th–8th cousins), but your close and moderate matches remain unaffected.


🌍 2. Ethnicity Estimates

  • Some effect — but minor overall:
    Missing SNPs mean less data for ancestry modeling. However, ethnicity algorithms are designed to tolerate some missingness (they use thousands of reference markers per region).

  • Still, a 1.8% missing rate might cause:

    • Slightly less confidence in small trace percentages.

    • Minor shifts in boundaries between similar populations (like between France and Germany, or between Indigenous Americas regions).

  • It won’t, however, cause a dramatic change (e.g., losing or gaining a whole continent).


🧩 3. Raw Data Uploads to Other Sites

If you upload your Ancestry file to GEDmatch, MyHeritage, or FamilyTreeDNA:

  • Some platforms will impute missing data (estimate it from neighboring SNPs).

  • Others will just treat those as blank, which can reduce overlap with other users’ files and slightly lower your match count.


In short:
✅ No major effect on close matches or main ethnicity regions.
⚠️ Possible small shifts in minor ethnicity regions or very distant matches.


Zero Zero, instead of a letters, means they could not determine your genotype. Apparently AncestryDNA treats no calls as wildcards, and assign a genotype. 

More Analysis from ChatGPT

I decided to ask ChatGPT to analyze my DNA for more ancestry and health related markers. I've checked the markers at 23andMe to make sure they were reported by ChatGPT correctly. ChatGPT seems highly accurate, but you should confirm if a marker is of particular importance to you. I've also uploaded to Gemini AI, but found they didn't always report the genotype correctly (meaning AG, TT, etc.). 


 It can only give very limited information regarding ethnicity. ChatGPT relies on only a few markers for ethnicity related characteristics, such as skin color and hair texture etc.. It doesn't have a reference panel. It would need a reference panel to provide more ancestral information. 

Health information was more detailed. It searches the web for information about your health related markers. The information it provided about my Thyroid disease could prove very helpful. According to one genetic marker I might have a hard time processing the thyroid replacement hormone and may need to change my treatment. Since I have continuing symptoms this might help? 

Ancestral information from ChatGPT:

What the markers show (straight facts from your file)

These are the genotypes I found in your file for the 11 AIMs we checked:

  • rs1426654 (SLC24A5) — AA — strongly European (light-skin associated)

  • rs16891982 (SLC45A2) — GG — European-associated allele

  • rs12913832 (HERC2/OCA2) — AA — European (eye-color) allele

  • rs3827760 (EDAR) — AGheterozygous for the East-Asian/Native-American-associated variant

  • rs260690 — AC

  • rs4891825 — AA

  • rs4798812 — AG

  • rs2065160 — AA

  • rs4988235 (LCT) — AG — lactase persistence allele present (typical in Europeans)

  • rs12203592 (IRF4) — CC

  • rs2814778 (DARC / Duffy) — TT — not the African Duffy-null allele

Short interpretation of those genotypes:

  • Strong overall European signal — multiple canonical European alleles are homozygous or present (SLC24A5, SLC45A2, HERC2, LCT). That fits a French / Northwestern European ancestry profile.

  • One heterozygous EDAR (rs3827760 AG) — this is the interesting bit. The derived EDAR allele is common in East Asian and Native American populations and rare in most Europeans. A single copy (AG) suggests a small contribution somewhere in your ancestry that carries that variant — consistent with limited Native American or East-Asian-related ancestry, which is plausible in some French-Canadian pedigrees (due to historical admixture) or could come from other branches.
    — Important: one heterozygous EDAR does not imply large East-Asian ancestry; it’s just an informative signal that merits follow-up.

  • No signals of significant recent African ancestry in the small marker set (rs2814778 not showing the African-null allele).

I definitely have Native American ancestry, so that information is correct. 

What ChatGPT Can Tell You About Your Raw DNA Data (information from ChatGPT on what it can do)

Beyond ancestry reports and trait predictions offered by these companies, AI tools like ChatGPT can help you interpret your data in fascinating ways.

1. Understanding Your Genetic Variants
Raw DNA data is essentially a long list of genetic variants, called SNPs (single nucleotide polymorphisms). Each SNP represents a small change in your DNA that can affect traits, ancestry, or disease risk. ChatGPT can help you understand what specific variants mean by explaining:

  • How a particular SNP has been associated with health traits in scientific studies

  • How common the variant is in different populations

  • Whether it might contribute to certain traits like eye color, lactose tolerance, or even caffeine metabolism

2. Health Insights (With Caution!)
While ChatGPT can summarize research about genetic variants, it cannot provide medical advice or diagnoses. However, it can:

  • Identify genetic markers associated with certain conditions, such as thyroid disorders, cardiovascular traits, or metabolic differences

  • Explain the current state of research on a gene or variant

  • Suggest areas you might discuss with a healthcare professional

It’s crucial to remember that genetics is only part of the story. Lifestyle, environment, and other factors often play a larger role than a single genetic variant.

3. Ancestry and Ethnicity Clues
Even though raw DNA files don’t come with polished ancestry reports, ChatGPT can help interpret markers linked to specific populations or regions. For example:

  • Identify SNPs more common in Native American, European, or African populations

  • Suggest possible regional connections based on genetic patterns

  • Help you explore historical migrations reflected in your DNA

4. Traits and Personal Characteristics
ChatGPT can summarize associations between your DNA and non-health traits, such as:

  • Hair color, eye color, and other physical features

  • Taste preferences or sensitivities (like bitter taste perception or lactose intolerance)

  • Sleep patterns or exercise response

5. Research and Curiosity
Your raw DNA is a powerful tool for personal exploration. ChatGPT can guide you in:

  • Researching genes or variants you find interesting

  • Comparing your genetic profile to published studies

  • Exploring ancestry in a broader historical and geographical context

6. Limitations
It’s important to approach AI interpretations carefully:

  • Not all genetic associations are strong or clinically relevant

  • ChatGPT relies on publicly available research and cannot access private medical records

  • Always consult a healthcare professional for medical concerns

Conclusion

ChatGPT can turn your raw DNA data into a source of insight and curiosity. From ancestry clues to understanding your genetic markers, AI provides a way to explore your DNA beyond what consumer testing companies offer. While it doesn’t replace professional medical advice, it opens doors to learning more about yourself, your heritage, and the science of genetics.

Tuesday, October 14, 2025

AncestryDNA Ethnicity Update 2025: Lessons from My Cousins’ Results

 Why My AncestryDNA Ethnicity Results Are Less Helpful Than My Family’s

My AncestryDNA ethnicity results have always been the least helpful of anyone in my family. I believe it’s because I’m more ethnically mixed than most of my relatives. I have some less common ancestry from Burgenland and Nicaragua, along with more typical Western European roots.

I do think it’s interesting to see how accurate AncestryDNA’s regional breakdowns have become. I know my ancestry going back at least a couple hundred years, and so far, I haven’t found any non-paternity events. I’m sure if I traced the lines further back, I’d eventually run into a few, which is why I expect the ethnicity estimates to reflect more recent ancestry. If they reflect ancestry from farther back, it’s harder for me to confirm whether it's accurate or not since I don’t have the documentation to support it. This is especially true when it comes to Irish genealogy — I can only trace those lines back a couple hundred years with records.

For someone who’s adopted, these ethnicity estimates can provide real clues about their heritage. They can also be useful for the descendants of enslaved people. If the estimates continue to improve in accuracy over time, they may become even more valuable for people in those situations. Personally, I already have a good understanding of my ethnic background, so these results aren’t critical for me.

In my experience, looking at the family trees of close DNA matches actually gives a clearer picture of your ethnic origins. These matches can sometimes pinpoint your ancestors’ exact towns or villages. You can look for recurring place names among your matches to help narrow things down. AncestryDNA Journeys is another tool that can point you in the right direction. Journeys are actually more useful for recent ancestry.

Paternal First Cousins Comparisons

So, why are my first cousin’s DNA ethnicity results more helpful than mine?

His results pinpoint the exact area where our paternal great-grandmother, Helen Mullen Mason’s, recent ancestors came from. They were all from Connacht, Ireland — specifically around Pollaturick, Milltown, Tuam, Co. Galway — for several generations before her birth. Unfortunately, I haven’t been able to trace this branch of the family past 1800. Our Irish surnames include Mullen, McQualter, Huvane/Huane, and Rattigan. I do know that the Huvane/Huane name goes back quite far in the Kilvine area of Co. Mayo.

My paternal first cousin has 9% Connacht Irish ethnicity, which perfectly matches what we know about our great-grandmother Helen Mullen Mason’s background.



I have 2% ancestry from Connacht, along with a mix from other regions, and 9% from Leinster. That doesn’t seem to make sense. Perhaps I had ancestors from there further back in time, but since my cousins’ results are so different, I doubt it. We don’t share the exact same DNA segments, and even so, the Leinster percentage seems questionable. 


It appears some of my Leinster is from my mother's side, but the highest percentage is from my father's side. Ireland being a small island it's likely that I have ancestry that traces back to many areas in Ireland. The Leinster result isn't very useful if someone is looking for hints regarding more recent Irish ancestry. 

Here are the percentages for all 3 of my paternal first cousins who tested at AncestryDNA. All have a higher percentage of Connacht DNA. I have 2%. If I didn't know my ancestral origins already I might guess my Irish ancestors were recently from Leinster, Ireland. 


Why Does AncestryDNA Show Zero French or French Canadian DNA—Even Though We Have a French Canadian Great-Great-Grandfather?

One of the biggest surprises in our DNA results is that none of us show any French or French Canadian ethnicity on AncestryDNA, even though we know — without a doubt — that our great-great-grandfather was French Canadian. This isn’t a case of a misidentified ancestor or a non-paternity event. We've confirmed the connection through solid DNA matches on that line.

The first result in the row is my result, the second is my mother's. She has no French recent ancestry. The next three rows belong to my paternal first cousins. 



French Canada has been added to regions, specifically Quebec. None of my family members have this admixture. 


So, why is AncestryDNA missing this?

1. Ethnicity estimates are not exact

AncestryDNA ethnicity results are just that — estimates. They're based on comparing your DNA to reference panels of modern populations, and that means there’s always some margin of error or generalization. Labels like “French” or “Quebec” may not reflect exactly how your DNA is inherited across generations.

2. DNA inheritance is random

You inherit about 3.125% of your DNA from a single great-great-grandparent. Because of random recombination, it’s entirely possible that the small DNA segments you received from that French Canadian ancestor don’t contain markers that Ancestry identifies as French or Quebec-related. That doesn't mean the ancestry isn't real — it just didn't get passed down in a way that Ancestry’s algorithms detect.

3. Quebec is now a defined region — but you may not meet the threshold

AncestryDNA now includes a defined region for Quebec, which is a great improvement. However, not everyone with French Canadian ancestry will be assigned this region, especially if that ancestry comes from just one great-great-grandparent. The region tends to show up more clearly for people with multiple lines going back to Quebec, or more recent Quebecois ancestry. A single ancestor four generations back may simply not leave enough detectable DNA for Ancestry to flag the region.

4. French and French Canadian DNA overlaps with nearby populations

Genetically, French ancestry — especially from the early settlers of Quebec — often overlaps with populations in Belgium, Northern France, Western Germany, and even parts of the British Isles. French Canadian settlers came from various regions of France, many of which share DNA markers with other Western European populations, making it harder to isolate them as distinctly “French” in ethnicity reports.

5. Reference populations still have limitations

Even with the new Quebec region, Ancestry's ethnicity algorithm is only as good as its reference populations and sampling. While the database is growing, it may still lack the resolution to detect subtle French Canadian DNA in someone with limited recent ancestry from that region — especially if it's diluted over several generations.


So, while our great-great-grandfather was clearly French Canadian — and we have both the family records and DNA matches to support it — AncestryDNA’s ethnicity estimate doesn’t reflect that, and that’s not unusual. Ethnicity results can be interesting, but they’re only a rough sketch. For deeper insights, looking at shared matches, family trees, and regional migration histories usually tells a more accurate story.

The ethnicity results are much better than they were when these tests were first offered. They still aren't accurate enough to paint a complete picture of your ethnic roots. 




Tuesday, June 17, 2025

Finding Winifred Mullen: A Long-Lost Aunt, a Boy from Mackinac Island, and an Unexpected Family Connection


My father Robert Kapple standing behind his sister around 1940 (about when Patrick Chapman visited). On right, you can see how much information I've collected on Winifred Mullen since earlier this month.

Winifred Mullen Found!

For years, my Great-Great Aunt Winifred Mullen was just a name in old Irish census records—born around 1892, living in Ireland in 1901 and 1911. I assumed she had stayed there her whole life, unlike her siblings—my great-grandmother Helen Mullen-Mason, her sister Bridget (who became Sister Mary Kathleen), and their brother Thomas—who all immigrated to Chicago (their brother Michael Mullen remained in Ireland until his death).

But Winifred had a secret: she, too, had crossed the Atlantic.

The Ellis Island Breakthrough

After years of searching Irish records for a marriage or death certificate, I finally found her—on a passenger list. In 1913, she arrived at Ellis Island, her name misspelled so badly that previous searches had missed her. Like her siblings, she was headed to Chicago.

But then… she vanished again.

The Chicago Clue

Frustrated, I turned to Ancestry family trees—usually a last resort due to unsourced information—and there she was: Winifred Mullen, married to Henry Chapman in Chicago in 1929. No parents listed, but the details fit (but her married name, Chapman, led me to find more records):

  • By 1930, they were living on Mackinac Island, Michigan.

  • The 1930 Census gave her year of immigration as 1913 (oddly her marriage record didn't come up in the searches until I included the name Chapman)

  • They had a son, Patrick Henry Chapman, born in 1931 

And then I remembered something my father had mentioned decades ago…

The Little Boy from Mackinac Island

My father, born in 1933, once recalled a childhood visit from a young relative—a quiet boy from Mackinac Island, around his age. The boy, he said, was Native American, which made him wonder if we had Indigenous ancestry.

The family thought the boy hadn’t enjoyed his visit—until they heard he couldn’t stop talking about it when he got home.

Now, the pieces fell into place: That boy was Patrick Chapman, Winifred’s son.

The Métis Connection

Further research revealed that Henry Chapman’s family had Métis roots—descendants of French-Canadian fur traders and Ojibwe women on Mackinac Island. Patrick, as a tribal member, carried that heritage.

But our link to him wasn’t through his father—it was through his Irish mother, my long-lost aunt.

A Bittersweet Discovery

Winifred died of breast cancer in her 50s. Patrick passed in 2002, and his only daughter is also gone. I’ll never hear his memories of visiting Chicago or what he thought of my father’s family.

But this journey taught me something: Family stories hold hidden truths. A passing comment about a "Native American cousin" led me to an entire branch of my family tree—one that tied Ireland, Chicago, and Mackinac Island together.

If you’ve ever hit a brick wall in genealogy, don’t give up. Sometimes the answer lies in:

  • Misspelled names (check every variation!)

  • Unexpected locations (who knew Winifred went to Mackinac Island?)

  • Family stories (even vague ones can hold clues)

  • Indexing and Transcription Errors (errors in indexing can affect search results)

And if you have ties to the Chapman family of Mackinac Island, I’d love to hear from you. There are still more stories to uncover, and photos.

Tuesday, June 3, 2025

What Ever Happened to Winifred Mullen? Dealing with early 20th century Irish and American records

 

My Great-Grandmother Helen Mullen-Mason, Bridget (Sister Mary Kathleen), and Thomas, but not Winifred? 

 I've been researching my Irish roots for a quarter of a century. My great-grandmother Helen Mullen-Mason immigrated to Chicago with cousins in the 1890's. Her sister Bridget (Sister Mary Kathleen), and brother Thomas, followed a few years later. That left two siblings in Ireland, after 1911. Brother Michael inherited the family land and remained in Ireland until his death. What happened to their sister Winifred has been a mystery? I have not been able to find a marriage or death record for her in Ireland. Also, oddly, I have not found a civil registration birth record for her either? The other children all have civil registration birth records, even though they are wrong because the family registered the births late and wanted to avoid paying a fine. I've done lots of research in Ellis Island records and did not find her. Winifred has been difficult. 

Why have I researched the siblings of my great-grandmother? For one I'm curious about their lives and their relationship to my grandmother Helen. Did they keep in touch? Yes, they did. My father knew Bridget Mullen (Sister Mary Kathleen), his great-aunt. Apparently my grandmother, and her siblings kept in touch with Thomas, their uncle. I have not found any evidence they kept in touch with Winifred, however. Another reason to research these relatives is I have been looking for photos of my great-great grandparents who lived until the 1930s and I believe there must be photos. Photos, letters, or any information about them would be appreciated. 

The first mention I had of Winifred was in Bridget, Sister Mary Kathleen's, eulogy. When the 1901 and 1911 censuses for Ireland became available I found her on those censuses too. That's all I had about her. I decided to devote last Saturday to researching her. I asked the Gemini AI chatbot for advice on where to search to see if it came up with any new record groups? It did suggest writing to Sister Mary Kathleen's order, the BVM archives, to see if there is any information about her sister Winifred in her file. I wasn't going to get any information from them on Saturday. The AI also suggested getting the exact date of birth for Winifred. Unfortunately, 1892 is missing from the church register book. She should be somewhere in the civil registration books, since legally they were required to register her birth.

Old Irish civil registration records are available at the irishgenealogy.ie website. It's very difficult to find someone at this site because the search provides a limited sounds like search for surnames. If first names and surnames are badly misspelled you can't find them. When you click on a record you get a pdf digitized copy. You can't turn pages and look through entire years to find misspelled names. The only search parameters are generally the first and last names of the individual the record is for. Parents' names and other names on the documents generally are not searchable for birth records. There are boxes to search for parents, but those names were generally never indexed. The page before the digitized record generally had null for parents names. The only way to narrow your search is by the large civil registration districts. 

What you can do is  slowly search through the records, page by page. You can do that if you know the approximate year of birth, then you click on every name that comes up. Each digitized page has ten birth records on it. The problem with that is it will often involve opening, and closing over 100 pages. You'll be opening the same pages over and over again because there are a number of people on each civil registration page. After doing that for a while I decided to use the census records to find other people who were born around 1892 in the townland, and neighboring townlands, to pull up pages for the more specific area of Dunmore (I tried searching on Dunmore, but could only search on the larger Tuam area). I never found her birth record. Being able to narrow searches would be helpful. I have given up on finding her in these records for now. 



I moved on to searching at Ancestry.com. I decided to go through different record sets looking for her. I've been searching passenger lists for years and didn't expect anything to come from that search. I was wrong! Two Winifred Mullens immigrated from Ireland to Chicago in 1913. Researching in Irish records I found that Winifred wasn't a very common surname in the area of western Ireland, where my family is from. Two Winifred Mullens immigrating to the same city, the same year is very unusual, given that uncommon first name. 

Was one of them my Winifred Mullen? Yes, it was! Why hadn't I found this record earlier? Her surname was badly misspelled in the index. The place she was from was the townland of Pollaturick, County Galway, Ireland. Of course Pollaturick would have been difficult to spell, and generally my family would instead give the name of the nearest town, which was Milltown. Of course Pollaturick wasn't spelled correctly on the passenger list either. Now I know she was still alive in 1913, and came out to Chicago to visit the family or settle there? 





I was very excited to find something else about Winifred! I later realized I forgot to copy the second page of the passenger list. I went back to the passenger list search. I searched using the same names I used before. Only one Winifred showed up, and it wasn't mine. I guess the search doesn't bring her up everytime. There are glitches with the Ancestry.com search, and could be why it didn't come up for me before. I was able to use my search history to find it. 

Armed with this information I searched for a marriage, death, or any other record for Winifred Mullen. I found a marriage record from the 1920s for Winifred Mullen. I hoped this was my relative. I also saw that there was an obituary for her in the 1974 Chicago Tribune. The obituary index at Ancestry.com said she had a sister named Kate. I immediately guessed it was the wrong Winifred because she didn't have a sister named Kate. Then I thought maybe this is referring to Sister Mary Kathleen, and Kate was a nickname? There was also a last name, but it seemed to just be a jumble of letters.  I went to Newspapers.com to see if I could find the entire obituary. They didn't have the Chicago Tribune for 1974. I eventually ended up on the Newsbank, HeritageHub site. It didn't have a digitized copy of the newspaper, but it did have a full transcript of the obituary.  

After reading the entire obituary I realized this person was the other Winifred Mullen who immigrated to Chicago in 1913. She indeed had a sister Kate who did marry, so was not a Nun. She had brothers named Michael and Patrick. My Winifred had brothers named Thomas and Michael. 

Below you can see Ancestry.com had Kate's last name indexed as Onanagner, instead of  Gallagher. The obituary was probably indexed using AI. 








HeritageHub saved me the time and effort of trying to get a copy of the 1974 Chicago Tribune. I suppose the information at Ancestry.com was from AI. It would have been nice if Ancestry had a more accurate, and complete index for newspaper obituaries. Only one relative's name was on it, when there were also two other siblings in the actual obituary. 

I have not been able to find anything further on my relative Winifred Mullen, so far. The fact there were two in Chicago, who immigrated in 1913, and were born around 1892 makes it difficult. They were also around the same height. My Winifred was 5ft 6 inches, and the other Winifred 5ft 7inches. After the other Winifred marries it becomes easy to separate them. I will keep trying to find more about this mystery relative, as I wait for the 1926 Census for Ireland to become available which will provide more family information.  







Saturday, March 8, 2025

RootsTech 2025 My Virtual Summary/ The good and disappointing announcements

 RootsTech 2025

 I did enjoy attending RootsTech virtually, however, there were aspects I didn't like. RootsTech 2025 revealed a clear trend: genealogy is becoming increasingly pay-to-play. The relentless push for premium subscriptions and upgrades to access basic tree and DNA tools is deeply concerning. While innovation is essential, it shouldn't come at the expense of accessibility.

FamilySearch remains a vital resource, offering free access to a wealth of original records and tools for building and sharing family trees. Their dedication to providing digitized microfilms, not just transcriptions, is a testament to their commitment to accessible genealogy.

Unfortunately, the same cannot be said for many other major genealogy platforms. The constant barrage of requests for subscription upgrades and DNA kit add-ons feels exploitative. It raises serious questions about the future of genealogy research, especially for those who cannot afford these escalating costs.

Is genealogy becoming a hobby reserved for the financially privileged? 

The main topics and announcements covered at RootsTech 2025:

  1. AI was featured in many presentations. Using AI to do research and transcribe records. 
  2. MyHeritage and 23andMe had presentations about their new Ancient Origins features available to subscribers. 
  3. Ancestry.com announced the rollout of a new tree feature allowing users to add FAN club members to be added to their trees. You can build networks of people your ancestors were associated with in some way. You have to have a subscription, and also subscribe to Pro Tools. 
  4. One of the best announcements was that AncestryDNA will be adding a cluster chart feature. It's like the clusters at MyHeritage grouping matches who likely relate through common ancestors
  5. Ancestry also introduced a tool to allow you to attach DNA matches' family lines to your tree.  
  6. MyHeritage announced a Cousin Finder search feature, which uses trees to find cousins. You can find it under discoveries. I found 196 cousins using this search. 
My favorite presentations: 
  1. FamilySearch Full Text Search Is Even Better by Robert Raymond. Full Text search is my favorite feature at FamilySearch. I hadn't searched there lately. When I heard many additional records have been added I searched there again and found some new information. Really wonderful!   
  2. Learn Genealogy Faster and Easier with Artificial Intelligence Mark Thompson. I'm interested in learning more about AI for genealogy so I watched several of these kinds of presentations. This one was very good! 
  3. How Just One Match Can Help You Find Missing Ancestors in 2025 Diahan Southard. All of her presentations are excellent!
  4. How Does DNA Fit into Your Family Story? Identity. Connection. Answers. Diahan Southard 
  5. Combine Multiple DNA Test Types (How atDNA+YDNA+mtDNA+atDNA=More Answers!) Diahan Southard
  6. Telling Stories with DNA: Case Studies from PBS' Finding Your Roots CeCe Moore 
  7. Ethical Standards in the Genealogy Community: Town Hall Discussion CeCe Moore, Lynn Broderick, Dr. Amy Harris
  8. On the Record: An Introduction to Civil Registration in Ireland Daniel Loftus
  9. Funeral Traditions in the Victorian Era Cathy Wallace
  10. What the Heck is a Centimorgan?  Kelli Bergheimer
  11. AI Genealogy: Year Two: The 2025 Outlook: Where We’ve Been, Where We’re At, Where We’re Headed Steve Little

Court Records Find Beaver County, Pennsylvania

One of the discoveries I made with the full text search was sad, but I'm glad to confirm some family information. It also helps with a timeline for when my ancestor Mary Kurta-Kappel returned to Austria. 

My father Robert Kapple told me his paternal grandfather, Frank Kappel, had a bad temper. My father's family lived with his paternal grandparents until his parents bought their own house. My father remembered Mary Kurta-Kappel was hysterical when she heard my grandparents, and their children, were moving out of the house. She said her husband, Frank, would kill her. (My father was born in 1933, and I believe the family moved out about 1940.) 

I found two court records, for Beaver County, Pennsylvania, using FamilySearch full text search, after listening to the presentation, which confirmed Mary was right to be frightened. 

This court record from 1908 states Frank had hit Mary with his fist and drove her out of their house. She dropped the charges likely because she had a number of children by 1908 and needed financial support. 


At the same time Frank drove his wife Mary out of the house he also drove his children out. There were 6 children under the age of 10 as this record states. They were thrown out of the house at night. The Humane Agent brought charges against Frank on behalf of the children. 


Frank was fined hundreds of dollars. I'm surprised that, working as a laborer, he had enough money to pay the fines. 

Soon after these incidents Mary took her children back to Austria, where her sister was living. She was likely pregnant at the time of the assault. My grandfather Rudolph Kapple was born 9 months later while the family was in Austria. 

Mary returned with her children to America in August 1910, two years after the abuse incidents in August of 1908 (however Joseph Kappel was left in Austria with family and didn't join the family in Chicago until 1923). Why she decided to come back I don't know? She probably couldn't afford to support 7 children herself. 

Here is Mary, with some of her children, in Austria probably in 1909. 


I also found a few other family deeds using the full text search at FamilySearch. I'm very happy with the additions to this Labs project. You can find the full text search under FamilySearch Labs. 

Personally I'm sticking with searching free records and only buying anything if I need it to solve a brickwall.

Rootstech 2025 has been productive for me! 

Sunday, February 2, 2025

New MyHeritage Ethnicity Results Update/ (and yes I do work with matches also)

 I've been looking forward to seeing the MyHeritage ethnicity results update for years now, and it finally came out a few days ago.

(The tools for working with DNA matches are much better at MyHeritage than at Ancestry. I really wish more people would test with them.)

As for my own update, I’d say it’s not particularly impressive for me, though it’s better for my mother and an aunt. Since I’m more ethnically mixed and farther removed from immigrant ancestors, my results tend to be less accurate than theirs.

In fact, my updated results are pretty close to earlier estimates from companies like Ancestry, FamilyTree DNA, and 23andMe. They still miss some of my grandfather’s ethnic contribution, which should account for about a quarter of my genes. He was of Austro-Hungarian ancestry, but I suspect the discrepancy is because most people testing with them are Western European. Additionally, I’m a quarter Hispanic, which isn’t as well represented either. Germanic and Eastern European should combine for around 25%, but instead they only add up to 12.7%. My Hispanic-related ethnicities should also total about 25%, but they’re listed as 16.5%. Still, the Hispanic predictions are better than the Austro-Hungarian ones representing my paternal grandfather.





Their genetic groups better reflect my ancestry, as you see below. 


My aunt’s updated ethnicity results are much more accurate. Her ancestry breaks down as 50% Austro-Hungarian, 25% Irish, 12.5% British Isles mix, and 12.5% French Canadian on paper. Her Irish percentage of 26.7% is spot on, considering her maternal grandmother was from Ireland. I figure 44.7% represents her Austro-Hungarian father—still a bit off from Rudolph Kapple’s genetic contribution but closer.



After adding up the numbers, I have to say the MyHeritage ethnicity update isn’t as bad as it initially seemed to me. If they had a larger database, which they should based on the quality of their product, the ethnicity results would be even more accurate.