Tuesday, 31 March 2020

Tree completeness/genetic confirmation

Covid-19 has a lot to answer for but one good thing is that we can work around the unexpected limitations imposed.
The DNA Discovery 2020 tour of New Zealand by Blaine Bettinger and Angie Bush could no longer take place, but thanks to the organisers, and of course the presenters, who are in widely different timezones from us, and each other, the 2 day set of lectures were changed to online sessions.
From a participant viewpoint, everything worked smoothly and well - thank you all, Blaine, Angie, Michelle, Fiona and Paul.

Blaine's talk on new tools for DNA analysis, reminded me I was going to check out the new DNAPainter haplogroup functionality (see eight ways you can use ancestral trees at DNAPainter) and also to see if I'd made any progress on both my tree completeness, particularly which levels had now been validated by genetic matches.

Posts such as these encourage me to actually review progress, and what I'm trying to achieve.
It's good to have snapshots of where I was in a point in time when I get discouraged about lack of progress - which is, I admit, rare as this is all so much fun.

Spoiler alert:
I've apparently made some progress 😀


The haplogroup progress will be a separate blogpost, as will the exploration of my supposed French-German ancestry (Yes Blaine, you inspired me to give biogeographical estimates another look and reminded me I was vaguely curious when I first looked at the segments 23andme identified years ago - it will be interesting to see what FTDNA comes up with in their promised equivalent).
On this page:

Tree completeness

For reminders of why names on trees and a few DNA matches is just a small part of the store front window of tree completeness read Blaine's blog post https://thegeneticgenealogist.com/2016/12/17/the-dna-era-of-genealogy/
Also Judy Russell's (thelegalgenealogist) on Filling in the Blanks

I've had a GEDCOM on DNAPainter since the beta tree functionality was announced, which means that two of my DNA-fuelled breakthroughs were already included in the tree.
The Lazy Vicar where the DNA leads/research had added parents and grandparents to my 2* great grandmother, so that was another branch complete to 4* greats.
I do so love a mystery or three likewise had added a whole 2 more ancestors to the level of 3* great grandparents, which of course only leaves more brickwalls to solve.

But what had not been included, given it only occurred in the last month was the review of my 3*great grandmother Jemima Parker's origins (Sarah not Charlotte; DNA strikes again refers).
So although she already had parents in my tree the tree has shrunk by one at that level (Jemima being the baseborn child of Sarah Parker) but grown by 4 more for her newfound mother's ancestry
Admittedly the link could still be somewhere back up from Jemima's husband, William Clinton about whom we know nothing beyond his marriage information and name on the baptismal records for their children.
Checking which is where the haplogroup confirmation can come to the fore - IF I can find a mitochondrial DNA descendant of Sarah Parker. There's only one found to date on paper, but not yet in person, and his surname is rather difficult for web searching - Wedding. But that's for the forthcoming haplogroup functionality post.

So with those added into my tree clicking on the highlighted Tree completeness link
gave me this table of how many ancestors I had at each level, and tells me about any ancestors who appear more than once (Pedigree collapse)

as you can see my tree completeness starts falling apart at the 4* great grandparent level and rapidly drops off after that.
One assumption at the 4th great level, not included here, would add 1 more at 4th, and 2 more at each of 5th and 6th great grandparent levels, all of which have interesting clusters of atDNA matches under active investigation - but given the stats on DNA inheritance at these levels the odds are fairly slim for making quick progress. I have however included the names in the surname hints/clues spot on DNAPainter.

The average amount of atDNA an individual can inherit from a specific ancestor halves as each generation passes, which does not necessarily mean that you will have inherited any from your btickwall one - but another cousin might have.
Encourage cousins to test, particularly those in the older generations still around on any of the lines of active investigation.
They beat your odds of not inheriting the crucial bit of DNA that will demolish the brickwall.

Check out the tables on the ISOGG WIki for the stats
- https://isogg.org/wiki/Autosomal_DNA_statistics and
- the theoretical probabilities https://isogg.org/wiki/Cousin_statistics
So it could take me some time to confirm the above assumption given the clusters of matches share a couple who married in 1738 and are, hopefully, my 5*great grandparents.
That's 8 generations (starting with me as the 1st) from whom, anyone in my generation can only expect to inherit 0.024% from one of those 5*great grandparents.
The likelihood of sharing significant matches with a 6th cousin is small, but possible.
Take comfort from the reported relationships with detectable matches at 6C and beyond in the Shared cMs project v4 also on DNAPainter.

Genetic confirmation

My first attempt at documenting which ancestors I'd confirmed at the time I started to pop this in DNAPainter resulted in the full fan tree view:

 shrinking to only this many being confirmed genetic ancestors:

which rather reflects my getting side-tracked rather than lack of confirmations, or I hope it does.

I admit I tend to work more on documenting the confirmations on WikiTree which pops dinky little symbols against each ancestor marked as "confirmed by DNA" (see WikiTree help for the WikiTree definition of / requirements for Confirmed by DNA status).
But the presentation there is by no means as compact - nor as colourful :)

is just a portion of my WikiTree "compact" tree view (see https://www.WikiTree.com/treewidget/Henderson-2297/5 for all lines of the 8 generations available.)
This portion shows the newly changed parentage for Jemima Parker and reminds me I've not documented the confirmation back to Sarah Parker a fully triangulated DNA match thanks to my maternal aunts, a 2nd cousin, and a 4th cousin who all share with several descendants of Sarah Parker's later marriage.
Additional shared match networks exist with other descendants and also corroborate that another family is also connected somehow which research is yet to unearth - see my Clinton - Parker brickwall page on WkiTree

So what does the genetic tree look like after some quick work?

My use of confirmation differs between the two sites (DNAPainter and WikiTree).
On the DNAPainter tree I use "Genetic ancestor" check box

to show the Ancestral Couple that a match and I share out to 4th cousins, even where I may not have fully confirmed which of the couple any DNA segments come from.
This differs from WikiTree confirmation in that the latter is strict on identification only of the individual relationship that has been "proven", particularly once you get beyond 3rd cousins.
The same quandary of which to show exists when painting your chromosome maps.
You have to decide whether to show the ancestral couple you share, or the person one generation below that the DNA match has now confirmed gave you that bit of DNA.
Depending on whose map I'm working on, I tend to use which specific ancestor gave me/them that bit of DNA - but it sometimes gives me the feeling I'm not making much progress back up the generations given it stops one earlier. 😒

Thank you to Jonny Perl for DNAPainter, and the DNA Discovery 2020 team for making me actually stop exploring matches and do an overdue stock take.

Sunday, 16 February 2020

Sarah, not Charlotte; DNA strikes again

DNA has done it again.

I've come up with a new theory about the parentage of my 3*great grandmother Jemima Parker thanks to a set of DNA matches to my maternal 2nd cousin.

This has led to a review of known, and believed, information about Jemima and a fairly easily reached conclusion that a Jemima baptised 1788 in Saffron Walden to a Sarah Parker is a much better fit than the previously identified 1793 baptism in Braintree Essex to parents William Parker and Charlotte Parkerson.

It really does pay to review early, well all, research occasionally.
A previous researcher had determined the 1793 baptism was correct and was unquestioned by me in my early days of working on this family back in the 1990s.
Jemima's death certificate however does say she was born in Saffron Walden - and that her father was a William Tyrrell. Well I suppose he may well still be, jury out on that one.
As a year, 1788 does fit the scant other information we have slightly better than 1793 but there's not much in that.
What is not so easily proven is that this Sarah Parker is one and the same as the one later marrying James Tyrrell in 1793, also Saffron Walden but I am comfortable with this assumption (for now).

If correct, the relationships of the DNA matches to the descendants of three lines of descent from Sarah Parker married to James Tyrrell (1793 Saffron Walden) range from 34 cMs to a descendant (Half 4C1R) of daughter Elizabeth; 8cMs to a descendant (Half 5C) via daughter Mary Ann; and 7cMs to a descendant of son William (a Half 5C3R).

Updates have been published for my web pages and WikiTree, other trees etc will follow in due course.
As to proving this, the circumstantial evidence is the Saffron Walden connection between the families and the timing of Jemima's baptism vs mother Sarah's subsequent marriage.

More DNA tested descendants showing up as matches on any of the sites would help.

So if you are a DNA tested descendant of any of those listed on my Clinton / Parker brickwall WikiTree page, please do make sure you are on all the sites accepting transfers (MyHeritage, FamliyTreeDNA, GEDmatch) even if you tested elsewhere or at only one of these.
Transfers are free, although more tools are available at each site for a small fee,
Having your DNA on any of these increases the chances of turning up matches to cousins who have not tested where you have.
Be aware that the more distant the relationship beyond 3rd cousins (itself only 95% odds of a mach being detected) the chances of a DNA match between any two specific people gets increasingly smaller, no matter what the paper trail.
But you may still match someone else who is - and holds the clue.

Don't know how to do transfers?
Check out Roberta Estes' blogpost for instructions

If you need help for transfers, please get back to me, I'm happy to help.

Saturday, 11 January 2020

New year, new updates

2020 is starting off a lot better than 2019.
Not that I make resolutions for a New Year as I'll only break them within the first few days (no chocolate? pay more attention to housework?) I do tend to check what hasn't been updated lately.

My gateway web pages (LornaHen) the related more detailed family website (BigBrother) and the basic BDM snapshots of the deceased rellies and their connections (RootsWeb Worldconnect replacement) have all had a refresh.
It's been a while and several links had become obsolete or unworkable (WorldConnect in particular, why did they change it? It's not for the better.).
Research has also moved on and new connections made. You all know how genealogy and DNA goes!

Perhaps the revitalization from a New Year will also extend to the long overdue updates to DNASurnames, and the Fairbairn and Runciman sites as there have definitely been new findings in the first two in particular with BigY results arriving.

It's not that there hasn't been any activity, just that what has been updated has mostly been tidied up on the FamilySearch tree and WikiTree, as have a lot of DNA connection investigations.

For the more intriguing of the latter I've started using WikiTree custom pages to keep track of the connected families and DNA conclusions as a result of the investigations.

WikiTree allows you to create custom pages for anything you wish and operates a category system to link profiles of interest.
Most of these are eg by place, surname, one name studies etc which have a monitored hierarchy but can also include Personal Categories which you link to yourself as the main grouping.

If a person on WikiTree has been associated with a category a tiny almost invisible link will appear at the top of their profile page
clicking on the link takes you to the bottom of the page where there is a list of categories associated with the person, eg my fledgling investigation into what I hope will be new Turnbull matches
One resolution I have made, is to review my chromosome maps and hopefully show progress since last year. 

Saturday, 16 March 2019

I do so love a mystery or three....

... particularly when it looks like a new lead may have appeared.
My one remaining brickwall mystery at the 2* great grandparent level is William Austin (this page will get updated hopefully soon as it reflects an old theory that is looking to be supplanted by the following, even though the mother's suggested name of Caroline is still possible, just not the one currently shown.

Courtesy of DNA matching it is looking rather like that mystery may now have some chinks in it.
Which all look to have some fairly stubborn brickwalls behind them!

On this page:

  • The DNA
  • The Research
  • The Theory
  • The Update


I have to caution myself that as yet I do NOT know that the match is definitely from the line of my match's paternal grandmother's ancestry that appears to be "of interest", but it is looking increasingly possible.

I've ruled out it being from her maternal side thanks to an overlapping sizable segment of DNA she has with a known maternal 1st cousin and my aunt who does not match the 1st cousin, so different ancestral sides for that bit of DNA (thank you GEDMatch for the ability to compare across sites for those uploaded to there).

Her closest match on MyHeritage looks to be from her paternal grandfather's side and none of my kits there match him/her.
But nor does their DNA overlap "ours" at any point, so inconclusive.
Added circumstantial evidence however is that it appears my match may have a crossover point between paternal grandparents on chromosome 10, as one of our matches abuts a segment from the above closest match.

(chr 10 DNA Painter segment to be inserted here, but seems a bit reluctant to do so)

Regardless of where the DNA eventually proves to be on her ancestry it is still a new clue to someone on mine as this is our closest unidentified match and we are all very near the top of her match list, so there is definitely a connection somewhere yet to be identified.

The research

Enough of the DNA, what about the research that has to go hand in hand with it?

As mysterious families go the family of William AUSTIN and Emma Parker CLINTON have thrown a few curlies into my research over the years.

Daughter Caroline Emma AUSTIN's name and origins being the first one.

I've used her as a case study in a study session or two as to the traps official records may set us.

My first official cert. purchased was for her marriage in 1882 to Matthias ROWE, reasoning that as the first she would be the source of the info for, and should know her parents.
She didn't even know her name (or age), let alone her father!
Caroline Emma Austin GIBSON, daughter of William Austin GIBSON, laborer, and Emma CLINTON.
By the time she died, the informant for her barely legible death cert. provided the information that Emma ROWE was the daughter of __ AUSTIN, labourer, (step father John GIBSON) and mother Emma GIBSON maiden name unknown.

The murk didn't clear much further with her mother's 31st Dec 1855 marriage cert. to Capt John GIBSON in Melbourne.
Emma CLINTON obligingly stated she was single and had no children from any former marriage (Caroline Emma was already 1 by this point).
Back in Feb 1853 before emigrating to Victoria Australia on the "Roxburgh Castle", Emma CLINTON of full age, daughter of William CLINTON had married William AUSTIN, porter, no father shown.
Beyond this cert. and the baptism of Caroline Emma in Feb 1854, Parish of St James, Melbourne to William and Emma AUSTIN, I have no further sightings of William AUSTIN.
No 1841, no 1851 census, no death.

Perhaps that's a "had", no further sightings?

Keep an eye on the developing thread on RootsChat re
William Edward Sinclair Fitz AUSTEN b. 1869 Mile End Old Town
son of William Mallars Fitz AUSTEN / AUSTIN (and looking increasingly to be aka William Mallars FITZAUSTEN) and wife Sarah BIGGS formerly BLAKE
(historical RootsChat thread continued in the above)

The main mysteries here being William Mallars Fitz AUSTEN / AUSTIN / FITZAUSTIN's whereabouts prior to the 1869 marriage and birth of son William Edward Sinclair Fitz AUSTEN and his parents.

See also the matching family being developed /summarized on WikiTree
Also on the FamilySearch tree

Both of these latter two are collaborative, free to use trees.
Feel free to add any research you have to either, with sources.

The Theory

My current theory from all of this is that William, like Emma, had re-married, but William put a bit more distance between his old and new lives by heading back to England, whereas Emma merely crossed the ditch to New Zealand.

If true and can be proven this makes the DNA match that prompted all this a half 3rd cousin to myself (and 1st cousins), and half 2nd cousin once removed to my aunts, and the amounts of DNA shared fits well with the observed stats on the DNAPainter interface to the shared cMs project

Given I'd never found a death for William AUSTIN I've always called Emma my bigamous 2xgreats granny, much to my father's disgust.
This makes it look like what was sauce for the goose was also sauce for the gander.

When I rang home many years ago to update my parents with the above find re Emma's marriage cert. to Capt John GIBSON saying she was single, with no children, my Dad tut tutted
"oh, you don't want to find out that sort of thing" he said.
"That's half the fun Dad why would you care?
It was back in 1855 - and Mum's side of the family"

The Update (Jun 2019)

Additional weight for my above theory is developing in the form of a new match to all of us on MyHeritage.
A 2nd cousin once removed to our initial match, so shared common ancestors are William Edward AUSTEN and Alice Maud CARTER.
With a fully triangulated segment on chromosome 10 for my family and theirs so we all still have a common ancestor who gave us that bit of DNA at least.
Now to rule out the CARTER ancestry and avoid confirmation bias!
We await a test from Ancestry which should show patterns for this latter.

Aug 2019

Ancestry results now in do not show any shared CARTER ancestry matches with the few kits we are able to comare on Ancestry. In addition, the sole close-ish CARTER cousin that is easily identifiable as also being on GEDmatch, and has no matches to any of the relevant AUSTIN kits, so AUSTIN is looking even better.

Not a lot of tested cousins on these lines for any of us makes it harder to draw solid conclusions here but I think I'm now convinced I know what happened to my 2*great grandfather William AUSTIN, after his arrival in Melbourne, so that's my last brickwall at the 2*great level declared demolished.

Saturday, 29 December 2018

and for the downside....

For the second time on FamilySearch the same person has (re)merged my paternal grandfather with his namesake born in the same month and year, despite my collaborative note - which is now yet more detailed as to why NOT to merge them.

They were born a fortnight apart, one registered in Oamaru, the other in Palmerston North, ie separate islands of New Zealand.
The separate NZ Historical BDM index entries show their different parents had been added to a note on each, but that didn't deter this person.
Yes on the surface it looks like the one who married first may have moved islands and remarried but he didn't. He remarried a bit further north.
It doesn't help that the older fiche birth indexes for New Zealand don't make it easy to find them to add the place details to the exact birth dates and parents available from the Historical BDM index online at the Dept of Internal Affairs.
Both are indexed as Archifold Hinderson instead of Archibald Henderson!

This time I've added yet more research into the note on both profiles.
The North Islander has been placed safely up in Te Kuiti with his 2nd wife in the 1919 electoral roll, his father's obit has been added  as well, which places him in Cambridge (Waikato, North Island) whilst my granddad is recorded down in Woodhaugh, Dunedin in 1919 with his first and only wife Agnes later moving to Roxburgh, Otago (South Island).
I do know that he once ventured north, at least as far as Wellington, on his honeymoon, but he did not already have a son by then, at least not Cyril.

I've also sent the chap a message via FamilySearch to PLEASE STOP merging these Archibalds.

WikiTree has also been updated with a disambiguation note on Granddad, with the other chap now also added there, he is after all nearly family - his son married my mother's 2nd cousin!
But there are other things I'd rather have been researching last night in the time spent doing the additional research that should, hopefully STOP the "helpful" person from merging them again.

Thankfully such persistance is rare, and an apology has just been received.

This does not detract from the overall benefit of sharing tree research.
We do not own our ancestors but we do owe it to them to be remembered accurately.

Such trees do tend to get better over time as duplications are reduced, research notes added etc.

Just the other day I was re-examining the scant tree of a DNA match on Ancestry wondering where his missing link to my DAWE family was and this time round, someone had added exactly the link I was missing.
The FamilySearch tree has now been extended to link him back to the rest of his ancestors already there.
This allowed me to breakthrough that mystery and connect in a newfound 3rd cousin thanks to both DNA and the collaborative tree at FamilySearch.

So mark your key people to be "Watched" so you can keep an eye on changes, you never know what you will learn, or who you will "meet" as a result.

Collaborative trees - the upside

Many of you may already have realised I'm a great fan of collaborative trees, WikiTree and FamilySearch in particular.

The downsides (enthusiastic newbies connecting up families with the scantest of evidence, or agin the evidence without a thought) are far far outweighed by the benefits of sharing.

Examples of both hit home strongly last night. (See next post for yesterday's example of the downside.)

For the huge plus side, my weekly Watch list email from FamilySearch highlighted that someone had added both a previously unknown child to the family of my 4x great grandparents Archibald Fairbain and Alison Crosser, and a baptism source for my 3x great grandfather Walter. (FamilySearch / WikiTree)

Which latter in particular is a bit embarrassing.

After all my research into Fairbairns you'd think I'd have gone back and rechecked my early work looking for the gaps that could now be filled.

There he is, baptized just across the border in Norham.
Which fits his 1841 census entry saying he was born in England, even if he did recant on that by 1851.

It's hardly an out of the way place for them to have been.
Courtesy of google we can see that it was 4 miles between Norham and Swinton where by the time the 1851 census came round Walter said he was born, having forgotten that 10 years earlier he'd said "England".

Unfortunately, there appear to only be the two baptisms there, no marriage for Archibald and Alison to fill that record gap, but also none for the presumed eldest son Archibald to contradict that he is the one baptized in Whitsome & Hilton, Berwickshire in 1783. :)
There seems to be only one other Fairbairn baptizing his children around this timeframe (1760 to 1800), a Thomas, with a William and an Ann.

Thank you to the person updating the Fairbairn tree that brought this to my attention.
Updates will ensue on WikiTree (done) and my Webpages

Monday, 17 December 2018

Mapped chromosomes down to 12cMs

Here's my 2018 end of year snapshot overall chromosome map, down to 12cM segments, by assigned Grandparent to give me as good a picture as I can as at now - both for ease of reference and as a measure of progress at the end of 2019.

It was a bit easier than I thought it might be to update the chromosome maps from yesterday down from the 15cMs used for those to the 12cMs minimum segment size used for the below. I had apparently been keeping up with the new smaller matches after all.

Gained a whole 6% more mapped DNA but it took over 450 more segments (from 12 to 15cMs) to add that 6%.
Dec 2018 Chromosome Map, by Grandparent, of 12+cM segments - DNAPainter
Still way too many yellow/orange/brown mysteries in there, hopefully disguising good clues to the brickwalls.

FWIW a table of the numbers of mapped segments per Side and Grandparent, split by 12-15cMs and over 15cMs is thrown in for contemplation.

Side GrdParent 12-15 GT15 Total Result
A Paternal? or Maternal? Side A 4 6 10
B Paternal? or Maternal? Side B 1 6 7
M Ern ANDREWS b. 1881 (TNK) 60 103 163

Ern ANDREWS b. 1881 (TNK) assumed 25 24 49

Honor ROWE b. 1886 (TNK) 43 144 187

Honor ROWE b. 1886 (TNK) assumed 6 8 14

Maternal - branch unknown 86 143 229

Maternal assumed - branch unknown 6 15 21
P Agnes Manson DAVIDSON b. 1885 (OTG) 49 64 113

Agnes Manson DAVIDSON b. 1885 (OTG) assumed 7 3 10

Archibald HENDERSON b. 1883 (OTG) 106 176 282

Archibald HENDERSON b. 1883 (OTG) assumed 20 22 42

Paternal - branch unknown 51 29 80

Paternal assumed - branch unknown 12 12 24
Total Result
476 755 1231

39% 61% 100%
The data comes from the main testing companies, FamilyTreeDNA, MyHeritage, 23andme, ie the ones who actually supply the segment level detail necessary (ie excludes Ancestry who refuses too). With many testers only on their original company it is not always possible to compare matches with known relatives at the right spots for accurate assignments, hence the "assumed" and "unknown" categories.
Do please consider transferring your DNA file to GEDMatch where cross company comparisons may be made!