Greg's English roots

The subject of yesterday's post was the information about ethnicity said to be derivable from a person's DNA, mine in this case.

Today I thought I would look at the ethnicity conclusions extracted from my husband Greg's DNA data to see whether these are consistent with what we know independently about his family tree: to see how Greg's DNA ethnicity compares with his documented descent. I have enough tested matches between his known cousins and people with whom Greg shares DNA to be reasonably sure his paper-trail tree matches his genetic tree for several generations back.

I know the birthplaces of all of Greg’s great-great grandparents. Twelve were born in England, two in Ireland, one in New Zealand of English parents, and the other in Australia of English parents. Most of his English forebears were from the south of England. There is no hint in what we know of their surnames, religion, or occupations to suggest that Greg's great-great grandparents were recent migrants from outside the region.

AncestryDNA reports Greg’s genetic ancestry as 100% European:

•  78% Great Britain 
•  11% Ireland 
•  7% Europe West 
•  3% Iberian Peninsula 
•  1% Europe East

Greg's ethnicity estimate as predicted by AncestryDNA 22 July 2017 (click to enlarge)

Greg’s mother, who believed that some of her Cornish ancestors had come from Spain, would have been very interested in this. The ethnicity results seem to suggest that there was indeed an Iberian connection of some sort (though not necessarily through Cornwall).

MyHeritage also reports Greg’s ethnicity as 100% European:

•  North and West Europe 96.4% 
•  Irish, Scottish, and Welsh 46.0% 
•  North and West European 29.5% 
•  English 16.5% 
•  Scandinavian 4.4% 
•  South Europe 2.6% 
•  Iberian 2.6% 
•  Ashkenazi Jewish 1.0%

Greg's ethnicity estimate from MyHeritage as at 23 July 2017

AncestryDNA has recently added a new feature, which it calls 'Genetic Communities'. Ancestry predicts at the 95% confidence level that Greg belongs to the 'Southern English Genetic Community'. This certainly matches his family tree. 71 people with whom Greg shares DNA, including Greg’s first cousin, several second cousins, and known third cousins, have also been linked to the Southern English Genetic Community.

Greg's connection to the Southern English Genetic Community

AncestryDNA attempts to provide support for its notion of a 'Southern English Genetic Community' with a brief history of the region, noting that in the early nineteenth century London was the largest city in the world, that many people emigrated from London and southern England to Australia, New Zealand, Canada, and the U.S.A.

The Ancestry website links people in Greg’s pedigree to the 'Southern English Genetic Community' through their birthplaces, broadly indicated on a large-scale map.

screenshot showing Greg's pedigree and associated birthplaces overlaid with the Southern English community data for the early 19th century

screenshot showing Greg's family tree data combined with the Southern English genetic community pedigree. Many of Greg's forebears from Southern England did indeed migrate to Australia in the 1850s and siblings or cousins migrated to America and Canada (Immigration in the 1850s was by sea).

I think that when it is placed in context with genetic relatives and historical events, the 'Genetic Community' interpretation of family history specific to our family tree is impressive and convincing. It agrees with what I have found in my own family history research.

Greg, however, is being difficult. He says his ethnicity is dinky-di Aussie, and the best thing you could say about all these Southern Poms is that they were (unwittingly) proto-Australians.

Related posts

• Looking at my ethnicity as determined by DNA testing
• DNA testing results one year on

Further reading

• AncestryDNA have provided a white paper on ethnicity testing. It is dated October 2013. It talks about reference panels based on 3,000 samples from "individuals alive today who can trace their ancestry to a single geographic location." https://www.ancestry.com.au/cs/dna-help/ethnicity/whitepaper
• MyHeritage also provide information about how they calculate the ethnicity estimate: http://helpcenter.myheritage.com/DNA/Ethnicity-Estimate/. MyHeritage compares my DNA with the DNA of living people around the globe whose genetic ethnicity is known and refers to these people as the Founder Populations. MyHeritage claims that they have sampled the DNA of thousands of people and have a data set of more than 100 ethnicities and the ability to show ancestral roots with far greater resolution than any other DNA service.

Looking at my ethnicity as determined by DNA testing

Besides its use in genealogical research, DNA testing is also employed for a simpler purpose: people have their DNA analysed to find out where their family is from, to trace their ethnicity.

I thought it might be interesting to compare the ethnicity predictions from my own DNA test results with what I know about my family tree. I have enough tested matches between documented forbears and people with whom I share DNA to be reasonably sure I'm getting it right.

Map from MyHeritage showing events from my family tree overlaid with my ethnicity estimate. The events are arranged by country. (click on image to enlarge)

My mother is from Germany. On her side of the family five of my great great grandparents were born in Brandenburg, two in Baden-Württemberg and one in Schleswig-Holstein. Based on their occupations, surnames, and religion, I have no reason to believe their immediate ancestors were from other parts of Europe.

On my father’s side, five of my great great grandparents were born in Australia. Of my 64 paternal fifth great grandparents:

• 21 were English 33% or 16% of my DNA 
•  17 born Scotland 27% or 13% of my DNA 
•  18 from Ireland, though this may include some Anglo-Irish 28% or 14% of my DNA 
•  4 from Wales 6% or 3% of my DNA 
•  2 from the Isle of Man 3% or 1.5% of my DNA 
•  1 from France 1.5% or < 1% of my DNA
•  1 English but of French Huguenot extraction 1.5% or < 1% of my DNA

AncestryDNA predicts my ethnicity to be 100% European:

• 59% from Great Britain, which includes England, Scotland, Wales and the Isle of Man 
• 20% Europe East 
• 12% from Ireland 
• 4% from Finland / North-west Russia 
• 2% from Europe-West 
• 2% from Italy/Greece 
• <1% from the Iberian peninsula 

These results appear to underestimate the German contribution to my DNA mix. I'm not sure why. 

My genetic ancestry as reported by AncestryDNA on 22 July 2017. The orange  region is the genetic community "Munster Irish" which AncestryDNA calculates I "possibly" belong to. I will make that the subject of a separate blog post. (Click on image to enlarge).

MyHeritage also provides an ethnicity estimate. I uploaded the same data I had submitted to AncestryDNA to MyHeritage. The result: 100% European:

• North and West Europe 80.1% 
• English 73% 
• Irish, Scottish, Welsh 7.1% 
• East Europe 13.5% 
• East European 11% 
• Baltic 2.5% 
• South Europe 6.4% 
•  Italian 6.4% 

Ethnicity estimate from MyHeritage retrieved 22 July 2017

AncestryDNA and MyHeritage draw different conclusions from the same data. They agree I am 100% European, but differ considerably in the regional proportions.

I don’t think these ethnicity estimates are well defined. Where does 'southern' Europe start? What does 'Baltic' mean? And how reliable are the estimates? Where does the 6% Italian come from? Were some of my English forebears descended from the Roman invaders? What's the connection between ancient Rome and modern Italy?

To me there's nothing very interesting about the ethnicity results and there's certainly nothing in them at this time to guide my research.

Further reading

• AncestryDNA have provided a white paper on ethnicity testing. It is dated October 2013. It talks about reference panels based on 3,000 samples from "individuals alive today who can trace their ancestry to a single geographic location." https://www.ancestry.com.au/cs/dna-help/ethnicity/whitepaper
• MyHeritage also provide information about how they calculate the ethnicity estimate: http://helpcenter.myheritage.com/DNA/Ethnicity-Estimate/. MyHeritage compares my DNA with the DNA of living people around the globe whose genetic ethnicity is known and refers to these people as the Founder Populations. MyHeritage claims that they have sampled the DNA of thousands of people and have a data set of more than 100 ethnicities and the ability to show ancestral roots with far greater resolution than any other DNA service.

Related posts

• DNA testing results one year on

Discovering a DNA cousin through Wikitree and confirming with GedMatch

In the year since we received our autosomal DNA results I have explored various tools to help me keep track of the results: to help me find new cousins with whom we share DNA and to show how our document-based trees connect.

One of the tools I have used is wikitree.com. Wikitree is a single tree grown using traditional genealogical sources and DNA. It currently has more than 14.6 million profiles added by more than 430,000 genealogists. Wikitree has some useful DNA tools to help make sense of your DNA results.

screenshot of the front page of wikitree.com

I have added information about our direct forebears to Wikitree. I added each forebear manually, and this meant revisiting the facts and checking that I had reliable sources for dates, places and relationships. I did not merely upload a GEDCOM file. I wanted to review relationships and I did not want to create duplicates. Wikitree has a single profile for each person. It is important to remember that none of us own our ancestors and we need to work with others on the information we attach to each profile. In fact we benefit from working with other descendants.

In addition to adding your forebears to Wikitree, you can add details of the DNA tests you have taken. Wikitree adds the information that you have taken the test to all blood relatives within eight degrees of separation — up to sixth great-grandparents and out to third cousins. You don’t upload the contents of the DNA results, just the fact that you have taken the DNA test and information that will help potential matches find you in each testing company’s database.

Recently I was contacted by Simon Bass, a distant cousin, who has also been adding his forebears to Wikitree. (Note Simon reviewed this post as a draft and following publication and is happy for me to blog about this case study.)  Simon has found that he seems to have a DNA connection to my husband’s family. Both my husband Greg and Greg’s brother Dennis have had their DNA tested and added their information to Wikitree. Simon is descended from Elizabeth Gilbart née Huthnance (abt 1774-1847) and her husband John Gilbart (abt 1761-1837). When adding their daughter Catherine’s details to the tree, linking Catherine to Elizabeth and John who were already on the tree, Simon noticed that Wikitree had a section on DNA connections on the right hand side of the screen.

Wikitree profile of John Gilbert (abt 1761-1837) retrieved from https://www.wikitree.com/wiki/Gilbart-8 on 15 July 2017. (click to enlarge)

Simon wrote to me and, despite testing with different companies, we were able to compare our kits on GedMatch.com . Our biological cousinship was confirmed.

GedMatch.com provides DNA analysis tools for genealogists including tools for comparing your own DNA test results with those of other people in the GedMatch public database. To use these tools you must first upload your DNA test results to GedMatch. GedMatch accepts results from Ancestry.com, Family Tree DNA, MyHeritage, 23 and Me, and WeGene.

GedMatch one to one comparison of DNA shared by Simon Bass and Dennis Young. Simon and Dennis share DNA on two chromosomes. Based on the size of the match it is estimated that there are six generations to their most recent common ancestor. Their relationship based on our document-based family tree is 5th cousins which is six generations to the most common recent ancestor, the prediction based on shared DNA is in line with our genealogy.

Since this first exchange of information Simon has shared photos with me of his trip back to St Erth in Cornwall, where the Gilbart and Huthnance families came from. We have also exchanged notes on the emigration of various members of the Gilbart family to Australia.

Wikitree made it easy for us to see that our document-based trees connected and showed that we had taken DNA tests. Having uploaded our test results to GedMatch.com, we could compare test results and see if there was a likely biological connection.

If there had been no shared DNA it would not have disproved that there was a relationship. It just meant that the same segments of DNA had not been inherited by Simon and his cousins Greg and Dennis.

Simon, Greg and Dennis are 5th cousins since they share 4th great grandparents. There is a nearly 70% chance that any two fifth cousins will not share a detectable level of DNA. ( https://isogg.org/wiki/Cousin_statistics ) In fact Greg’s brother Dennis shares DNA with Simon but Greg does not.

Related posts

• Edwards family immigration on the Lysander arriving in the Port Phillip District in 1849
• DNA testing results one year on

DNA testing results one year on

To see what DNA testing might show about our family history, in June last year my husband Greg and I submitted samples of our DNA to Ancestry.com for autosomal DNA analysis. Greg also took a Y-DNA test through the Family Tree DNA company.

What did I hope to learn?

I had no particular questions about my side of the tree that I thought DNA tests might answer, though I did hope that I might find some of my relatives on my mother’s side who, I had been told, had emigrated from Germany to the USA.

My mother was born in Berlin. After WWII when there wasn't enough to eat, food parcels addressed to her family began to arrive from the United States, sent---she was told---by her American cousins.  When I talked to her about my DNA testing she was a little sceptical but she thought it would be interesting to trace her American relations.

On Greg’s side of the tree there were some dead ends that I hoped DNA analysis might provide a way around. In May 2016 I blogged about these in a post [Poor little chap], with a hypothesis about the parents of Greg’s great grandfather Henry Sullivan (1862-1943).

When our test results came back I quickly found some meaningful matches on Greg’s tree. Some of his cousins had tested too, and we were able to make links between our pre-DNA trees and the DNA information.

I started a spreadsheet to record these matches and, to confirm the links I had found, I sent messages through AncestryDNA's system to people who appeared to have matching results.

My spreadsheet has these fields:

• Who took the test - me, my husband, a cousin
• Match name and/ or Nickname (for Ancestry, GEDmatch, My Heritage, 23andMe alternate names)
• Company or companies testing
• point of contact – sometimes the kit is managed by someone other than the person who provided the sample
• Email address for match or contact
• Received – when I receive a message or email to help me remember and track messages
• Notes – always a handy field
• Line, if known, or grouping of tests- family line, sometimes I know the line from shared matches or other clues, otherwise I can group tests which share matches
• MRCA (Most Recent Common Ancestor) – I enter the 2 surnames of the couple I also use Ahnentafel numbers at the beginning of the field which are handy for sorting generations
• Cousin (cousinship) – eg: 3C or 4C-1R
• estimated relationship
• shared surnames
• linked tests (shared matches)
• Tree – copy in the URL of any tree or site for the match’s tree
• GEDmatch ID number of match
• Chromosome summary
• GedMatch Autosomal Comparison - copy and paste from the Gedmatch comparison (this can get a bit lengthy and detailed so I don't always include it)

A screenshot of the first 11 rows of my spreadsheet where I track matches. I sort the spreadsheet on most recent common ancestor and since the Ahnenetafel number is used to begin the field, the spreadsheet is sorted with more recent ancestors first. (click on image to enlarge)

I also used the 'Notes' field on AncestryDNA when I had contacted the match and added a star to matches so that I could see at a glance who I had contacted, when I contacted them, and the likely match.

A screenshot of an AncestryDNA match where our trees correspond and the most common recent ancestors are identified. There is a yellow star on the top left hand side of the screen which can be used to identify matches of interest. The notes field in the middle of the screen can be used to summarise contact, GedMatch numbers and likely shared ancestry.

A screenshot of a DNA match where our trees on ancestry do not intersect. In fact I have identified the most common recent ancestors through correspondence with another cousin and shared DNA. Although I messaged using the green "Send Message" button in the top right hand corner, the cousin did not reply. I also unsuccessfully used the brown "Contact" button on the top right of the screen for the Ancestry Profile - see screenshot below.

Initially my DNA matches meant nothing to me. There seemed to be little connection between my DNA matches and my document-based tree.

However, over the last year I have received replies from people we share DNA with, and have been able to compare our family trees and determine where they intersect. For many of these matches, I have been able to draw some conclusions as to our likely most recent common ancestors, particularly when several of us share the same DNA and have intersecting document-based trees. So far we have more than 30 matches where we can identify the most recent ancestors we share.

DNA has added to the evidence I have used to build and document my family tree. To date I have not found any DNA-based evidence which suggests my document-based tree is inaccurate. Unfortunately I have yet to find any cousins sharing DNA on my mother's side of the family.

Our family tree showing points on the tree where we have common ancestors with  cousins who  share DNA. (You can click to enlarge, but the tree has been reduced in size and individual forebears are not legible)

Related posts

• Poor little chap - considering what we might learn from testing our DNA
• DNA analysis: slow progress
• DNA analysis: taking the tree back two generations
• DNA analysis: matching the DNA results to the"paper tree"

The Legal Genealogist observes in her post of July 2 2017 That YDNA lament "you can't be matched to somebody who hasn't tested" and that there are good reasons for the lack of German DNA matches to date.

DNA analysis: matching the DNA results to the"paper tree"

My first DNA results from AncestryDNA were a surprise to me.I could find no matches to my family tree. How was I going to break the news to my parents that I had been swapped at birth?

I had DNA matches to other people's DNA results but no links to my tree. This seemed odd for I have been researching my tree for many years. I can identify all my 3rd great grandparents and a large number of my more remote forebears.

At present I have 98 pages of matches on AncestryDNA. At 50 matches per page this is just under 4,900 people who seem to share DNA with me and have tested with ancestry.com. Of these, 21 are estimated to be 4th cousins or closer, the closest being an estimated 3rd cousin who shares 127 centimorgans of genetic linkage across 8 segments; statistically, there is a high confidence we are related. The most distant match shares 6.0 centimorgans across 1 segment. (That match is linked to a private tree. I don't know how we might be related.)

AncestryDNA summary as at 4 February 2017, as of 5 February I have 21 matches calculated to be at the 4th cousin level or closer.

On my ancestryDNA summary page I now have 6 green leaves, which in the AncestryDNA system indicates shared DNA together with a connection in our trees. I can see our connection in 5 of these. The sixth has a private tree, so I cannot see the link. When my DNA matches were first made there were only two green leaves and both these were to my 7th great grandparents, Daniel Dana (1663-1749) and Naomi Cresswell (1670-1751), a connection too remote to be reliable. (We may connect differently on an as yet undocumented part of our tree.) Without a chromosome browser, which shows shared segments with other matches who have the same ancestry, I cannot be sure I have correctly identified our most recent common ancestors.

a shared DNA match where both our trees include common ancestors

My 6th great grandfather Richard Dana (1700-1772), portrait by John Singleton Copely painted about 1770. Portrait now in the collection of the Metropolitan Museum of Art

It is not surprising that the matches were to the Dana branch of my tree. The Dana were early emigrants to the United States. As the Dana family is well-documented I would expect members of it to show up in people's trees and to find connections on that branch. A chromosome browser would give me more confidence that the connections have been made reliably.


Of my top 21 matches, that is, matches predicted to be 4th cousins and closer, I have managed to identify the most common recent ancestor in four cases. Only one of my AncestryDNA green leaf hints is estimated to be a 4th cousin and it would appear she is actually a 7th cousin once removed. Even when there are only small trees that do not connect with my family tree I have had some success in working out the connections with my DNA matches .

Spreadsheet analysis of my matches who are estimated to be 4th cousins or closer (click to enlarge)

AY, whose kit is administered by JB, is estimated to be my 4th cousin with 40 centimorgans shared across 3 segments. The match shows there is no family tree linked to the kit. However, when viewing the match, although a family tree has not yet been linked to the DNA results there are trees created by JB that can be previewed. In this case there were 3 trees,  

• one with 27 people and no surnames I recognised and no apparent link to AY whose DNA was tested
• The second was similar, with only 5 names
• The third, although it had only 6 people, had a name I thought I recognised and seemed to have AY as the initial person. I searched my tree for the surname Yeates and found that in 1857 my 3rd great aunt, Dymphna Maria Cudmore (1836-1899) married Sydney van Butchell Yeates (1831-1918). 

screenshot of my DNA match with AY showing public but unlinked trees in the bottom left of the screen

AY and I are 3rd cousins once removed. Our most recent common ancestors are Daniel Michael Paul Cudmore (1811-1891) and his wife Mary Cudmore née Nihill (1811-1893).

More recently I had a shared match with AY. TH was estimated to be a 4th cousin and we share 20.7 centimorgans across 2 DNA segments. TH had no tree. However, because of the shared match I speculated that we might be connected on the Cudmore or Nihill lines.

shared matches are shown when you click on the middle tab; the green contact button is on the top-right of the screen but is not always effective as there is no email notification that a message has been sent

profile page showing brown contact button in top-right-hand of screen

In September 2016 I contacted TH using the green button on the top right hand side of  the DNA match screen but did not hear back. After reading that communication with DNA matches using the brown contact button on the top-right-hand side of a person's profile page, was likely to be more successful, I tried contacting TH by that method and providing my email address in the message. He replied by email almost immediately and confirmed a connection to the Nihill or Niall or Nihell family. We are 4th cousins once removed.

Daniel Michael Paul Cudmore (1811-1891), my 4th great grandfather, about 1865. Image from the State Library of South Australia reference B30912

Mrs Mary Cudmore née Nihill (1811-1893), my 4th great grandmother, about 1865. Image from the State Library of South Australia reference B 30913

If your research is like mine I have a few suggestions:

• Although a DNA match may show as having no tree, check if there is an unlinked public tree available to view. In my top 21 matches, 8 apparently had no tree, but in 4 cases there was an unconnected tree.
• A connection can be made with a tree of very few people. (I found two connections by following the descendants of a 3rd great aunt.)
• Shared matches give hints as to how other DNA matches might be connected.
• DNA matches might not be ancestry subscribers and, even if they are they might not log in often. Consider contacting them by using the brown contact button on their profile page, providing your own email address. Contacting by the brown button provides an email alert. There is no similar alert for contacts made by the green button.

DNA analysis: taking the tree back two generations

DNA: SNP model from Wikimedia Commons

Progress on our family tree using DNA evidence has been slow. Cousin matches are either with people where we have no idea how they fit into our tree, or with cousins whose place in the tree we already know. The second kind of matches are useful, however, because they help to predict where on our  tree those otherwise unlinked matches might belong.

When we received our first DNA results in July 2016 one of the first matches I contacted was A B. She was predicted to be Greg's 4th cousin with AncestryDNA stating Greg and AB shared 21.6 centimorgans across 2 DNA segments. She had a private tree, so I was unable to view what links there might be between our tree and hers. Her member profile gave no hint as to where in the world she was.
Later in the month AB replied and shared her private tree with me. Neither of us could see where the link was. We both uploaded our DNA kits to GedMatch.com, which confirmed the link, giving slightly more information than AncestryDNA had provided:

Comparing Kit A828918 (*G C Y) and Axxx (*AB)
Minimum threshold size to be included in total = 500 SNPsMismatch-bunching Limit = 250 SNPsMinimum segment cM to be included in total = 7.0 cM

Chr	Start Location	End Location	Centimorgans (cM)	SNPs
8	22,956,646	32,079,918	12.6	1,731
12	64,128,525	77,383,427	13.3	2,210

Largest segment = 13.3 cMTotal of segments > 7 cM = 25.9 cM2 matching segmentsEstimated number of generations to MRCA = 4.6
441334 SNPs used for this comparison.

At the time no other kits uploaded to GedMatch matched AB and Greg. AncestryDNA also showed no shared matches.

AncestryDNA offers a view of surnames and places of birth that two trees have in common. We noticed early on that there were a large number of places in Lincolnshire from AB's tree and some close to those on our family tree.

.
The orange markers are birthplaces on AB's tree. The blue from our tree, and the green are birthplaces appearing on both trees.

Because of the Lincolnshire birthplaces we looked at both Dawson and Plowright lines as possible connections but came to no conclusions.

In September AB looked again at her tree, focusing on her great grandfather John William Gibbons. AB had noticed that she and her father shared DNA with an AncestryDNA match, To2, and that Greg also shared DNA with To2, although not the same segments (hence not showing as a shared match).  AB found that her shared ancestors with her father and To2 were John Gibbons and Frances. Frances was possibly the daughter of Robert Atkin and Frances Smith.

Greg had no forebears with the surname Gibbons in his tree but looking at his DNA matches there were some matches who had Gibbons in their tree, in particular several had Rebecca Gibbons (1843-1897).

Rebecca was born in Moulton, Lincolnshire, the daughter of Thomas Gibbons. In 1866 she first married William  Noble Waite (1845-1879). They had five children and emigrated to the United States in the 1870s. William Waite died in Utah. Rebecca's second marriage was to Lemuel Sturtevant Leavitt (1827 - 1916) in 1882 in Utah, USA. I had noticed that several of Greg's DNA matches had Leavitt as one of the surnames.

Lemuel Leavitt was a Mormon pioneer who travelled overland to Utah at the age of 21 in 1849.

I used Wikitree to document some of my research on the branch and connect the branch to the wider single tree. Lemuel Leavitt was on the tree but I needed to create a profile for his wife Rebecca. Several ancestry trees included Rebecca but she was not well researched and facts were sometimes factually wrong, for example on one tree she was shown as being married to Lemuel Leavitt in 1850.

I found a possible Gibbons link to our tree, an 1826 marriage record in Horbling in Lincolnshire for Thomas Dawson. The spouse was Ann Gibbons, who lived there.

Greg's 4th great grandfather was Thomas Dawson (1775 Gunby, Lincolnshire - 1861 Bennington, Lincolnshire). He was married to an Ann.  I wondered if this was Ann Gibbons of Horbling.

AB identified Rebecca Gibbons Waite Leavitt in her tree, confirming a link to the trees for the descendants of Lemuel Leavitt and Rebecca Leavitt formerly Waite nee Gibbons with whom we shared DNA. These descendants had apparently not yet researched Rebecca's parents or origins.

In early November the  General Register  Office of the United Kingdom launched a new index of birth registrations. This revised index included mother's maiden names. From this, AB discovered the record of Betsy Dawson's birth in 1838. Bestsey was the youngest child of Thomas Dawson and Ann. The birth index showed the mother's maiden name Gibbons. Bestsey appeared n the 1841 census with her parents Thomas and Ann and sibling Isaac (1831-1872). Isaac was Greg's great great grandfather. There were two other siblings, Eliza and William. This gave us confidence that we had correctly identified the 1826 marriage of Thomas Dawson to Ann Gibbons as being the marriage of Isaac Dawson's parents.

Thomas Dawson married Ann Gibbons at Horbling which is 5 miles from Aslackby, where AB's Gibbons forebears lived.

We started to speculate. AB gave a surname to the wife of John Gibbons. She wrote:

I have added a surname to Frances (Atkin/s), wife of John Gibbons at the top of my tree. This is what is in many other people’s tree, and its true that the marriage dates seem to fit. HOWEVER - the reason why I am not sure is that the records for the parish of Aslackby for this time period have not yet been fully digitised, but are at the archives. I need to look at them anyway for my mum’s tree.

So having added the surname ATKINS it has thrown up some hints from other trees and in a couple there is a daughter ANN born 1801 Spalding, sister of my George and of Thomas - the ancestor of the LDS’s. I do feel that this would be about the right generational distance between our families.

I am not convinced about the accuracy of the online trees but it is worthy of further investigation. 

The next day on the lincstothepast website AB found a baptism for Ann on 16 December 1801, daughter of John and Frances Gibbonds, at St Mary and St Nicholas Church, in Spalding, 15 miles from Horbling .

This was the only ( I think) church in Spalding at this time, as st Pauls was built by my Quinton ancestors ( as labourers)  in mid / late 19thC. St Mary and St Nicholas was the church where John Gibbons and Frances Atkin were married.

The date seems to fit the age Ann Dawson when she died. In identifying a forebear I would not normally rely on such a slim connection but the DNA seems to be another piece of evidence, in particular the additional DNA matches to several descendants of Rebecca Gibbons.

In conclusion, DNA is really just additional evidence, to be reviewed with documents and indexes. Given the DNA evidence I am reasonably confident we have identified the maiden name of Greg's great grandmother Ann Dawson (1801-1842) and we no know who her parents were: John Gibbons (1780-1840) and Frances Atkins (1772-1856). This means that Greg and AB are 5th cousins, within the range predicted by AncestryDNA and GedMatch.

simplified tree, click to enlarge

DNA analysis: slow progress

My husband Greg and I had our DNA analysed by ancestry.com in May 2016. The results became available in July. Over the last six months I have spent many hours trying to make sense of the results and contacting DNA cousins.

As a way of learning more about our family history DNA analysis has not been very useful.

At the time of testing I put forward three brick walls the DNA testing might help me get past:

• Henry Sullivan (1862-1943), a great grandfather of my husband, was admitted as a state ward in Geelong in 1866 about the age of four. His parents had deserted him. 
• George Young (1826-1890), from Liverpool, a great great grandfather of my husband, provided no information about his parents. 
• Caroline Clarke (1835-1879), wife of George Young and a great great grandmother of my husband, was not specific about her birth and I have been unable to trace her parents, perhaps John Clark(e) and Hannah Sline. She said she was born about 1835 in Tumut. But she also said she was born in Sydney. 

I have made no progress on the history of George Young or  his wife Caroline Clarke.

DNA testing has provided some evidence that the hypothesis I explored for Henry Sullivan's family in my blog post of May 2016 is correct. It adds to the body of evidence which includes newspaper reports, an inquest, state ward records and birth, marriage and death records, all of which point to a connection.

MBH, a descendant of Henry's probable half-brother, has tested. She matches Greg, his first cousin Barry, and a second cousin of Greg's, LB, also descended from Henry Sullivan. My research suggests that MBH is Greg's third cousin,. They share a common great great grandmother but are descended from children of that great great grandmother by different fathers. AncestryDNA predicted the relationship between MBH and Greg as 4th cousins with 21.4 centimorgans shared across two segments.

chart showing family connections of related DNA tests

Three of the test results have been uploaded to GedMatch so that we can look at how the DNA inheritance aligns. However, the DNA from the three samples does not match on the same segments. 

GedMatch shows the match between Greg and MBH as having only one segment on chromosome 1 of 23.4 centimorgans. It predicts their relationship as having the most recent common ancestor within 4.6 generations. By lowering GedMatch's default thresholds I find that the second segment is a match on chromosome 2 of 5.9 centimorgans, a size regarded as not genealogically significant. Lowering the threshold still did not reveal a matching segment shared by Greg with both MBH and LB.

Barry, a first cousin of Greg's, also on the Sullivan side, has tested. He has yet to upload to GedMatch. The tools provided by ancestryDNA show that he shares DNA with Greg, MBH and a third person, IB. Barry separately shares DNA with Greg and LB, that is, Barry and Greg inherited different segments that they share with MBH compared to the segments shared by LB and MBH.

Greg, Barry, and MBH share DNA with a fourth man, IB. IB has a private tree on ancestry.com. Unfortunately he has not responded to my several attempts to contact him. As far as I know he is not a descendant of Henry Sullivan. MBH is unaware of his being a cousin. IB is predicted to be a fourth cousin of Greg's. His tree (865 people) does not appear to contain the surnames Hughes or Sullivan. 

Although it is likely that the DNA matching is for the hypothesised Sullivan and Hughes connections, it is not certain. We don't have enough shared matches on a specific segment to be confident that the shared DNA demonstrates shared heredity.  Greg, LB and MBH have family trees that have been researched over considerable time and go back a number of generations. There don't seem to be any other likely connections between the family trees.

Ancestry.com's DNA database continues to expand rapidly with over 3 million participants to date. Future matches may well provide more definite confirmation of inherited DNA from Matilda Sullivan née Hughes, the mother of Eleazer and probable mother of Henry.

Related post

• Poor little chap

Accentuate the Positive 2016 Geneameme

I thought I would take up Jill Ball's challenge again and look back on my 2016 family history researches.

1.  I found some elusive ancestors, the parents and grandparents of my husband's great grandfather Henry Sullivan: Poor little chap

2.  A precious family photo: a cousin of Greg's showed us her family album. In it was a copy of a photo I had from another branch of the family. The presence of this photograph in her album helped to confirm the identity of the two teenagers. Y is for Young family photographs

3.  An ancestor's grave: I have yet to visit the grave in Geelong of my 4th great uncle William Dana but I have seen some photos. On my to-do list for 2017 Trove Tuesday: death of Captain W. A. P. Dana

4.  An important vital record: the 1863 birth of Ebenezer Henry Sullivan - see Poor little chap.

5.  A newly-found family member shared her DNA. Together through shared matches and by reviewing our family trees and birth, marriage, and death records we added a couple of generations to our family tree on my husband's side. This collaboration showed that you need a tree to start the matching process and that DNA alone is not enough. DNA, of course, helps to verify documents that by themselves would not necessarily have been enough to have taken the family tree back further. I need to write up my genetic genealogy experiences.

6.  A geneasurprise I received was eating a cake from a recipe my grandmother often used to use. It took me back decades: O is for Oma cooking from Dr Oetker's "Backen Macht Freude" [geneacake? geneafood?]

Schwarzwälder Kirschtorte made according to a recipe my grandmother followed

7.   I am pleased with all my 2016 blog posts. In researching each blog entry I learned a little more about my family history and was always inspired to keep researching.

8.   I made a new genimate who ... after some years of correspondence it was terrific to catch up with Greg's cousin Sharon

9.  A new piece of software I mastered was Sun charts from MyHeritage.

10. A social media tool I enjoyed using for genealogy was Facebook. I found the group Using DNA for Genealogy - Australia & NZ full of useful advice

11. In order to get the most out of DNA testing I have spent a lot of time during 2016 viewing webinars about genetic genealogy. I found the advice by Crista Cowan on communicating with cousins particularly useful: https://www.youtube.com/watch?v=qG2CiY_6c9Y .

16. I continue to get much useful material from the State Library of Victoria and the Public Record Office Victoria, both in North Melbourne and in Ballarat. These are enormously important repositories of relevant information. On line of course the National Library of Australia's Trove has made a great difference to family history research and I use it frequently.

17. A new genealogy/history book I enjoyed was The Conversations We Never Had by Jeffrey H. Konis. I need to write up my own review of this book. I also enjoyed High Seas & High Teas: Voyaging to Australia by Roslyn Russell.