Saturday, February 14, 2015

Adventures in genetic genealogy: my progress so far

It's been about a month since I gave up waiting and started "making something happen" on 23andme. I have 983 "DNA Relatives" there, and Gedmatch shows me my top 1000 matches. I've messaged all of the 23andme matches, with mixed results.

First, only about one-third of the cousins have a name or profile that I can see on 23andme. Those people I have messaged by name, and if they have surnames and/or localities listed, I often mention those if they sound interesting or familiar. Many people have nothing listed in their profile but a name. Those folks without a name can be messaged, but it has to be generic. A few of those nameless people have accepted contact, and some "sharing", which is what 23andme calls comparing genomes.

Of course sharing genomes is the whole reason I'm doing this, so it has been a lot of work just to get to starting line. For those who are not interested in doing this work to get matches, my advice is: use FTDNA or Ancestry. 23andme does not make it easy. Even when you have contact information and can use email, finding the common ancestor is work; although more fun than writing endless messages on a crappy message system.

After one month, there are still 203 outgoing invitations, and perhaps 500 messages to nameless persons still sitting in their 23andme inboxes. I have 203 people sharing with me; 15 of those are not matches, which happens when one person administers a group of kits and allows sharing on all of them. Realistically, I expect another 10 to 20 more sharing matches over the next year. On the upside, that is 200 more cousins!

Now the bad news. Perhaps 35 of those 200 are on gedmatch; and most of them already were there before I found them. I've talked only 5 or so people into uploading there. Of course I need to make another push and re-message *all* my sharing matches, whether they are in a shared-segment group or not. And even being on gedmatch is only the first step to finding that common ancestor.

I focused at first on those shared-segment groups because it seemed obvious to me that they would *want* to get onto gedmatch (and hopefully wikitree), but that doesn't seem to be the case. And really, all of my 23andme matches are part of a shared-segment group, even if there are no matches with us on gedmatch, which is highly unlikely. So if I want to play the numbers, I'm going to have to write to all those singletons, and those whom I've not messaged for a month.

After all, when I first sent in my kit, after I spent some time setting up a profile (only to see my tree disappear), it was nearly a year before I checked in again and started figuring out how to use the site. I assume that most people are about the same, and have no clue 1. how or why they should set up a profile and make it public, 2. where they can get and send messages, 3. why and how they should compare genomes, and 4. why they should care at all. Much less know what the next steps might be.

It is too bad that 23andme don't do any of the education of their users, but they seem content to get their money and then leave them alone. Maybe that makes business sense, because the only way they make money on people is when they buy more kits. Ignorance on the part of their customers perhaps makes them the most money?

Whatever. They have given me some matches, and some tools. I need to make the best of what I have. Cousins, cousins, and more cousins! On to Wikitree, research, and finding those common ancestors.

Labels: , , , , , , , , , ,

Wednesday, February 04, 2015

Genetic Genealogy, Why?

I've loved the search for genealogy and my family's history since my children were young. I wanted to understand my roots, and why some of the difficulties my family endured happened, and how those events changed all those who were touched by them.

I found young orphans, babies born before marriage, rape, and tragic death by fire among the more normal events. I've also found some family lines that are rather well-researched, which took away the challenge. For some years other projects took my time.

Then, the yDNA tests became affordable, and I bought an Family Tree DNA kit for my dad for Father's Day. Because those "beginner kits" don't tell you much, and have far too many meaningless matches, I kept paying for upgrades, all the way up to 67 markers. This took his high-quality matches down to under 10, but as we followed up on these matches, we realized that this line of Cowans came from Stirling, in Scotland. The recorded part of the family went to the Irish Plantations as merchants. The Sterling Cowans were wealthy and powerful. They established the Cowane Hospital in Stirling, which we were able to see on a trip with Dad, my sister Kimberly and me a few years after my mother's death. I have yet to find documents proving the link between my Selkirkshire Cowans and that powerful family, but my dad's yDNA prove that it is there.

Now, autosomal DNA is affordable, and much more powerful than they have ever been before. The segments of each chromosome which are sampled are only the parts where we humans show difference. Most of our DNA is identical. So I asked for a 23andme kit for Christmas, and when my husband bought one for me, spit and spit and spit until I filled up the test tube.

See and for more. This is what really got me going:

It took a few weeks before I got results, and I saw very little that meant anything to me. Every time I re-visited the site, I saw more that made no sense, until I ran out of patience, and started the attack. Now I'm really getting results, if by results you mean cousins. As described in detail in my previous post, I wrote to all my 23andme matches. And as 23andme emails me about new matches, you can be sure I will follow up with them.

The reason we want all these matches is not just social; it is triangulation, which is assembling known data, and using that to solve the unknown. Described here:

So, a few hundred messages are sitting in cousins' inboxes. Meanwhile, 150 of us are beginning the process of finding our common ancestor. Some of those cousins are in groups of two to ten matches on the same segment with me. I hope to get those singletons in a group one way or another.

While these match requests are trickling in, I've been flogging Gedmatch. If you are wondering how to get started, there is an exhaustive PDF about how to use gedmatch:

What to do on the site?­hat-to-do-at-the-gedmatch-site­/,­edmatch-a-wonderful-tool/, and for those of you with a bit of cash to spare to help out the site. I have a membership now and find it well worth the ten dollars per month.

My main use so far is to find more cousins, and as more folks from 23andme upload there, we are able to verify which segment groups on gedmatch match us. Now we've started including those Gedmatch folks in our group mails. I will end up emailing them all too. Gedmatch doesn't send out emails about new matches, but when you do a "one kit with many" search, the new kit numbers are green.

Of course, it is still a waiting game; send out emails, and then wait for the replies. Kitty Cooper recommended using Wikitree, so I began investigating that too. I have a gedcom at Rootsweb: I love Rootsweb, and I love WorldConnect, but aside from the Postems, the trees are not interactive, or connected up.

Wikitree is set up differently. The goal is to collaborate on the research about each ancestor, and link up to cousins along the way. That alone is wonderful, but it also integrates DNA testing results into the ancestral profiles. I've uploaded my gedcom, and checked each possible match, and now my upload will be examined by one of the helpful experts who help keep the site humming. Politeness, integrity and downright friendliness are built into Wikitree. I love it.

See for some detail, and Randy Seaver's blog about doing the actual gedcom work. Of course I will keep my own database in Gramps, but I really love the idea of public collaboration with cousins.

Labels: , , , , , , , , , , , , , , , , , ,