Running successful atDNA projects in Ireland

Panel discussion: Genetic Genealogy Ireland 2018

2:30 p.m. Saturday 20 October 2018

Dodder Suite 5A, Main and Industries Halls, Royal Dublin Society

WWW version:

YouTube versions:

TBA

This is my attempt to answer Maurice Gleeson's 24-part questionnaire to panel members in the allotted 8 minutes!

My FamilyTreeDNA projects

My first public project was the Clare Roots project:

established to coincide with Maurice Gleeson's November 2015 talk to the Clare Roots Society;
modelled on the project supported by the Ballymena branch of the North of Ireland Family History Society.

I am now administrator or co-administrator of the following public projects (member counts as of 17 October 2018):

geographically-based:

Clare Roots (co-administator Terry Fitzgerald, WA, USA; 1153 members; 678 FF)
Kerry Y-DNA Project (administrator John Hallissey, Waterford; 69 members; 41 FF)

surname-based:

Clancy (administrator Fergus Clancy, Dublin; 90 members; 57 FF)
Durkin/Durkan/Durcan (administrator Mike Durkin, NY, USA; 28 members; 13 FF)
Marrinan (administrators Cory Marinan, WI, USA; Cindy Wood, MI, USA; Greg Marrinan, CT, USA; 33 members; 20 FF) (This project has an associated Facebook group.)
McNamara (rescued from worldfamilies.net; 102 members; 50 FF)
O'Dea/O'Day/Dee (administrator James O Dea, Dublin; 96 members; 40 FF)

[There are inconsistencies in the count of FF kits on the Order Summary page and the Project Statistics page (above). The latter appears not to count any autosomal transfer kits. The former appears not to count Y-DNA or mtDNA kits which subsequently add an autosomal transfer. It is unclear whether each page is counting orders submitted or results received.]
I welcome members with Y-DNA, mitochondrial DNA and/or autosomal DNA results to all of these projects, provided that they qualify for membership on either geography or surname grounds as appropriate.

My standard e-mail invitation, sent, for example, to people who contact me because they match a kit I manage:

I am co-administrator of the FamilyTreeDNA Clare Roots project which you can join by going to https://www.familytreedna.com/sign-in?ReturnUrl=%2Fmy%2Fgroup-join%3Fgroup=ClareRootsSociety, logging in to your FamilyTreeDNA kit if necessary and following the prompts.
If you want to see how many of your Family Finder matches are in this project (or any project that you have joined), then just go to the Advanced Matches page at https://www.familytreedna.com/my/advanced-matches.aspx, login if necessary and tick the Family Finder checkbox and select the project in the "Show Matches For" dropdown.
In order to achieve the full benefits of project membership, you must take action to give the project administrators "Limited" or "Advanced" access to your kit. This can be done when you join or can be done later at https://www.familytreedna.com/my/project-preferences. For each project that you have joined, click the orange Edit button and if necessary select "Advanced" or "Limited" in place of "Minimum" from the relevant dropdowns, then click Accept and then click Confirm. If you belong to more than 10 projects, then they will be split across multiple pages.

The "About" page outlines both benefits of membership and responsibilities of membership.

The Clare Roots project probably has an emphasis on the old Kilrush Poor Law Union because of my involvement in the Kilrush & District Historical Society and my recruitment and swabbing campaign is also largely in that area (318 FTDNA kits processed).

The Clare Roots project has outgrown the 1000 individuals shown on the (private, unhyperlinkable) member information web page and the 500 lines available on the (public, hyperlinkable) Y-DNA results page.

A county is probably too big an area for an effective DNA project or genealogy database; the Poor Law Union, more recently known as Superintendent Registrar's District, is a more appropriate size; the wide diaspora of a single parish alone or even of a single townland could constitute a very useful and interesting project.

The public DNA results pages make no reference to autosomal DNA, and the Y-DNA Colorized Chart is the only one that I use.

Project members are encouraged to copy their Family Finder results to GEDmatch.

Those of us involved in geographically-based or surname-based genealogical projects have come up with informal ad hoc ways of using the other DNA websites for such projects.

AncestryDNA projects

Marrinan descendants: only 33 members of the FTDNA project, but over 105 GEDmatch DNA kits, with over 30 more on Ancestry but not on FamilyTreeDNA or GEDmatch.

143 people have shared AncestryDNA match lists with one AncestryDNA account, effectively an autosomal project with no tools or sophistication: requires

a tool to avoid clicking individually on all kits on the Ancestry dropdown to find all matches to existing project members;
a chromosome browser; and
triangulation tools.

GEDmatch projects

I manage 46 GEDmatch kits.

GEDmatch tag groups are its nearest equivalent to projects.

Shared tag groups not yet implemented.

No facility to reorder members of saved tag groups, e.g. by family branch or by location.

Marrinan results are displayed using GEDmatch tag groups, broken down geographically - see printout.

Offline projects

My personal genealogy database:

edited with Ancestral Quest on my Windows laptop
copied periodically to Reunion on my iPhone
copied periodically (excluding details of living individuals) to TNG behind a password on my website
started in 1987
135,445 individuals as of 18 October 2018
I have found 1,752 of those individuals in one or more online autosomal DNA databases
I export ancestors of GEDmatch kits to a GEDCOM file if the user wishes
I export ancestors, and 3C or closer also in DNA databases, of Family Finder kits to a GEDCOM file if the user wishes
Should the GEDCOM database include all one's research, or should one have multiple GEDCOM databases for different surnames or different geographical areas?
My philosophy is that I should never have the same individual in two different master databases.
But good software should be able to create a single-surname database, by exporting all end-of-line ancestors with a particular surname, plus all of their descendants, as a GEDCOM, for use in an autosomal surname project.

I also keep a Microsoft Excel spreadsheet listing whichever of the following details I can find for autosomal DNA kits of interest, to enable quick switching back and forth between DNA and genealogy databases:

known relative of my own?
if so, closest relationship
current surname
first name
birth name (if known) or "adopted"
GEDmatch kit number
RIN from my database, if connected
GEDmatch alias
DNA company
gender
e-mail address
Facebook profile
AncestryDNA profile
AncestryDNA tree number
AncestryDNA cfpid
Formula to create hyperlink to Ancestry pedigree chart

Useful tools

The most useful autosomal tool provided to FTDNA project administrators for private research is the Family Finder Illumina OmniExpress Matrix, the equivalent of GEDmatch's Autosomal DNA comparison matrix.

Example:

Facebook discussion about the Hehirs of Cloverhill in IGP's County Clare Ireland Genealogy Group.
Descendants of probably unrelated Hehirs from all over the county chimed in.
If Michael Brooks who m. 7 April 1901 in NY to Ellen Clancy was Michael Brooks b. 25 March 1873 in Fountain to John Brooks and Mary Hehir, then the first person in the sample matrix would be 2C1R to the second person and 3C1R to the other four.
In fact, Michael Brooks of Fountain m. 19 Feb 1901 in Kilnamona to Mary McKee and never emigrated.
The descendant of Michael the emigrant is not on GEDmatch, but is in the FTDNA Clare Roots project, along with five known relatives of Michael of Fountain.
The display to five decimal places conveys a spurious sense of accuracy.
Rounding down to 0.00000 is at variance with the FTDNA policy of otherwise counting half-identical regions of 1cM/500SNPs.
Administrators needs facility to import the known or hypothesised relationships from a GEDCOM file and scale the shared centiMorgans by the expected shared centiMorgans for the known or hypothesised relationship: when looking at different relationships in a single matrix, "percentage of expected centiMorgans" would be a much more meaningful scale that plain centiMorgans.

My favourite GEDmatch tool is the Tier 1 matching segment search.

Communication with project members

Project administrators should be able to anticipate and answer questions before new members have figured out what to ask.
Different novices will want to be pointed towards different parts of the online information (FAQs).
Project activity feed is more appropriate for broadcasting to members than mass e-mails.
E-mail is probably better for one-to-one advice to individuals.
Sometimes phone or Skype is the only way to persuade members to do something.
We meet visiting diaspora project members somewhere like the "genealogy office" in the corner of the bar in the Stella Maris Hotel in Kilkee.
For many projects, Facebook posts have supplanted the activity feed (and Ireland Reaching Out parish pages) as the easiest way to facilitate mass communication to a wide audience.

Major challenges

Unanticipated consequences of GDPR:

Before GDPR changes, administrators of multiple projects could switch quickly between kits in any of their projects via bookmarks;
Since GDPR changes, there are no longer the same benefits from having separate projects (or having a private project for people who don't fit anywhere else, of which the main benefit was avoiding the need to repeatedly log out of one kit and log in to another);
FTDNA's interpretation of GDPR is that mailto: links to the e-mail addresses of all project administrators must be displayed publicly on the website to spammers and non-FTDNA customers, where pre-GDPR e-mail addresses were available just to matches and optionally to project members;
my long-standing guidelines on e-mail etiquette forbid anyone from publishing my e-mail address online, in order to protect it from spammers.

It can be hard to persuade project members and others to:

answer e-mails or AncestryDNA or MyHeritage messages;
add pedigree charts;
add most distant known patrilineal and matrilineal ancestors (MDKAs);
upload to GEDmatch and other databases;
order Y-DNA upgrades;
etc.

My major successes

Adoption reunions;
O'Brien/O'Dea surname/DNA switch;
llinking branches together;
verifying or refuting the results of traditional research;
predicting a fourth cousin relationship purely on the basis of surname and townland, and subsequently finding a 0/67 Y-DNA match;
etc.

The future

Cuimhneamh an Chláir (Memories of Clare) slogan: 'When an old man dies, a library burns'.

We owe it to our descendants:

to record the memories, the oral traditions and the DNA of our older relatives, all of which often go back beyond the history can be found in surviving paper records;
to do unto our descendants as we wish that our ancestors had done unto us;
to leave our DNA to our descendants as we are trying to reconstruct that of our ancestors; and
to record and keep alive the history of local areas and families.

Those running DNA databases owe it to their stakeholders (and those participating in the databases owe it to fellow participants) to provide the best possible service, information and tools, as in the following wish list:

Wish list (cf. Challenges above)

Easier sharing and comparison across DNA databases.
Every DNA kit to have a public pedigree chart and use the full 50 characters to describe MDKAs.
Women to always use their maiden names (but not to call themselves Mrs Maiden-name!).
Better algorithms to identify ancestors who must be dead, instead of calling them "HIDDEN" or "Private".
Every member of every DNA database to have enough public information on ancestors that there is never a need for follow-up questions.
Separation of paternal and maternal autosomes.
Couple the DNA database to the genealogical database by embedding DNA kit identifiers in the GEDCOM, allowing more automation:
- Combine the GEDmatch matching segment search with FTDNA's Family Matching:
  For each half-identical region in the matching segment search output, indicate:
- Search my genealogy database to identify:
  - all end-of-line ancestors with a given surname;
  - variant surname spellings which might be included;
  - all descendants of these end-of-line individuals;
  - those of the descendants who are in the DNA databases;
  - their lines of descent from the end-of-line ancestor;
  then:

Acknowledgements

Thanks to Cindy Wood of the Marrinan project for many of the thoughts and suggestions above.