We have a brand new toy in our DNA sandbox today, thanks to Don Worth, a retired IT professional. I just love it when extremely talented people retire and we, in the genetic genealogy community, are the benefactors of their Act 2 evolution. Our volunteers make such a cumulative difference.
Drum Roll please.
Introducing…..the Autosomal DNA Segment Analyzer, or ADSA.
The name alone doesn’t make your heart skip beats, but the product will. This tool absolutely proves the adage that a picture is worth 1000 words.
Don described his new tool, which, by the way, is free and being hosted by Rob Warthen at www.dnagedcom.com, thus:
I created this tool in an attempt to put all the relevant information available that was needed to evaluate segment matches on a single, interactive web page. It relies on the three files for a single test kit that DNAgedcom.com collects from FamilyTreeDNA.com. These files include information about your matches, matching segment locations and sizes, and “in common with” (ICW) data. Using these files, the tool will construct a table for each chromosome which includes match and segment information as well as a visual graph of overlapping segments, juxtiposed with a customized, color-coded ICW matrix that will permit you to triangulate matching segments without having to look in multiple spreadsheets or on different screens in FamilyTreeDNA. Additional information, such as ancestral surnames, suggested relationship ranges, and matching segments and ICWs on other chromosomes is provided by hovering over match names or segments on the screen. Emails to persons you match may also be generated from the page. The web page produced by this program does not depend on any other files and may be saved as a stand-alone .html file that will function locally (or offline) in your browser. You can even email it to your matches as an attachment. You can play with a working sample output here.
Who wants to play with sample output? I wanted to jump right in. Word of caution…read the instructions FIRST, and pay attention, or you’ll wind up downloading your files twice. The instructions can be found here.
I can’t tell you how many times, when I’ve been working with matches, that I’ve wondered to myself, “How many other people match us on this segment?” For quite a while you could only download 5 people at a time, but now you can download the entire data file. I’m a visual person. To me, visually seeing is believing and the ADSA makes this process so much easier. Truly, a picture is worth 1000 words.
I knew right away there were three things I wanted to do, so I’m going to run through each one of the three by way of examples to illustrate what you can do with the power of this wonderfully visual tool. I’ve also anonymized the matches.
1. Clusters of matches.
I know I’ve told you that I’m mapping my DNA to ancestors. When I first saw Don’s output, I knew immediately that this tool would be invaluable for grouping people from the same ancestral lines.
Barbara, the second row, is my mother and her DNA in this equation is extremely useful. It helps me identify right away which side of my family a match comes from. If you don’t have a parent available, aunts, uncles, cousins, all help, especially cumulatively.
Before we begin working with the results, take a minute and just sit and look at the graphic below. These two clusters shown on this page, one near the top and the other at the bottom….they represent your ancestors. Two very different ones in this case. This may be the only way you’ll ever “see” them, by virtue of a group of their descendants DNA clustered together. A view through the keyhole of time provided by DNA. Isn’t it beautiful?
All of these results in this “cluster of matches” example are my matches. In other words, the file is mine and these are people who are matching me. You can see that this tool provides us with start and end segments, total cMs and SNPs, and e-mails, but the true power is in the visual representation of the ICW (in common with) matrix. The mapped segments are a nice touch too, and Don has listed these in progressive order, meaning from beginning to end of the segment (left to right.)
Look at this initial clustered group, shown enlarged below. The first individual matches me and mother on one pink segment, but matches me on two segments, a pink and a black. That means he’s from Mom’s side, or at least through one line, but probably somewhat distant since that one segment is his only match on any chromosome. Because he also matches me on a segment where he doesn’t match Mom, he could also be related to me on my father’s side, or maybe we had a misread error on the black segment when comparing to Mom’s DNA. It is the adjoining segment. In essence, there isn’t enough information to do much with this, except ask questions, so let’s move on to something more informative.
Beginning with the third person, the next grouping or cluster is entirely non-matching to mother, so this entire cluster is from my father’s side AND related to each other.
There are 6 solid matches here, and then they start to trail off to matches that aren’t so solid.
By flying over the match names with my cursor, I might be able to tell, based on their surnames, which line is being represented by this cluster of matches. If I already have a confirmed cousin match in the group, then the rest of the group can be loosely attributed to that line, or a contributing (wife) line. Unfortunately, in this case, I can’t tell other than it looks like it might be through Halifax County, VA. I do have an NPE there and some wives without surnames.
Let’s look on down this chromosome. There is another very solid cluster, also on my Dad’s side. In this second cluster, I have identified a solid cousin and I can tell you that this is a Crumley grouping. My common ancestor with my Crumley cousin is William Crumley born about 1765 in Frederick Co., Va. and who died about 1840 in Lee Co., Va. His wife is unknown, but we have her mitochondrial DNA. Now this doesn’t mean that everyone in this group will all have a Crumley ancestor, they may not. They may instead have a Mercer, a Brown, a Johnson or a Gilkey, all known wives’ surnames of Crumley men upstream of William Crumley. But someplace, there is a common ancestor who contributed quite a bit of chromosome 1 to a significant number of descendants, and at least two of them are Crumleys.
At first, I found it really odd that my mother had almost no matches with me on chromosome 1. Some of my mother’s ancestors came to the States later, from the Netherlands and from Germany. Many of these groups are under-represented in testing. However other ancestral groups have been here a long time, Acadians and Brethren Germans. My father’s Appalachian, meaning colonial, ancestors seem to have more descendants who have tested.
However, looking now at chromosome 9, we see something different.
The second person, Doris, doesn’t match Mom anyplace, so is obviously related through my father, but look at that next grouping.
I can tell you based on hovering over the matches name that this is an Acadian grouping. The Acadians are a very endogamous French-Canadian group, having passed the same DNA around for hundreds of years. Therefore, a grouping is likely to share a large amount of common DNA, and this one does.
Based on this, I can then label all of these various matches as “Acadian” if nothing more.
Within a cluster, if I can identify one common ancestor, I can attribute the entire large group to the same lineage. Be careful with smaller groups or just one or two rectangle matches. Those aren’t nearly as strong and just because I match 2 people on the same segment doesn’t mean they match each other. However, when you see large segments of people matching each other, you have an ancestral grouping of some sort. The challenge of course is to identify the group – but a breakthrough with one match means a likely breakthrough with the rest of them too, or at least another step in that direction.
2. Source of DNA
I have several cousins who match me on two or more distinct lines. This tool makes it easy, in some cases, to see which line the DNA on a particular chromosome comes from.
I have both Claxton (James Lee Claxton/Clarkson born c 1775-1815 and Sarah Cook of Hancock Co., TN) and Campbell (John Campbell b c 1772-1838 and Jane Dobkins born c 1780-1850/1860 of Claiborne Co., Tn.) ancestry. My cousins, Joy and William do too. In this case, you can see that Joy matches a Claxton (proven by Y DNA to be from our line) and so does William on the first green matching segment. The second green segment is not found in the Claxton match, so it could be Claxton and the Claxton cousin didn’t receive it, or it could be Campbell, but it’s one or the other because Joy, William and I all three carry this segment.
What this means is that the light green segments are Claxton segments, as are the fuchsia segments. The source of the darker green segment is unknown. It could be either Claxton or Campbell or a third common line that we don’t know about.
3. Untangling Those Darned Moores
I swear, the Moore family is going to be the death of me yet. It’s one of my long-standing, extremely difficult brick walls. It seems like every road of every county in Virginia and NC had one or more Moore families. It’s a very common name. To make thing worse, the early Moores were very prolific and they all named their children the same names, like James and William, generation after generation.
The earliest sign I can find of my particular Moore family is in Prince Edward County, Virginia when James Moore married Mary Rice (daughter of Joseph Rice and wife Rachel) in the early 1740s. By the 1770s, the family was living in Halifax County, Virginia and their children were marrying and having children of their own of course. They were some of the early Methodists with their son, the Reverend William Moore being a dissenting minister in Halifax County and his brothers Rice and Mackness Moore doing the same in Hawkins and Grainger County, TN. Another son, James, went to Surry Co., NC. We have confirmed this with a DNA descendant match.
We have the DNA of our Moore line proven on the Y side through multiple sons. At the Moore Worldwide DNA project, we are group 19. Now there are Moores all over the place in Halifax County. I know, because I’ve paid for about half of them to DNA test and there are several distinct lines – far more than I expected. Ironically, the Anderson Moore family who lived across the road from our James and then his son Rev. William, who raised the orphan Raleigh Moore, grandson of the Rev. William Moore, is NOT of the same Moore DNA line. Based on the interaction of these two families, one would think assuredly that they were, which raises questions. This Anderson Moore was the son of yet another James Moore, this one from Amelia County, VA., found in the large group 1 of the Moore project. If this is all just too confusing and too close for comfort for you, well, join the crowd and what we Moore descendants have been dealing with for a decade now.
This raises the question of why there are so few matches to our Moore line. Was our Moore line a “new Moore line,” born perhaps to a Moore daughter who gave the child her surname. However, the child of course would pass on the father’s Y chromosome, establishing a “new” Moore genetic line. I’m not saying that is what happened, just that it’s odd that there are so few matches to a clearly colonial Moore line out of Virginia. With only one exception, someone genealogically stuck in Kentucky, to date, all DNA matches are all descendants of our James. We do know that there was a William Moore, wife Margaret, living adjacent to James Moore in Prince Edward County but he and his wife sold out and moved on and are unaccounted for.
I’ve seen this same pattern with the Younger family line too, and sure enough, we did prove that these two different Y chromosome Younger families in fact do share a common ancestor.
So you can see why I get excited when I find anything at all, and I mean anything, about the Moore family line.
A Moore descendant of Raleigh, the orphan, has taken the autosomal Family Finder test, and he matched my cousin Buster, a known Moore descendant, and also another Cumberland Gap region researcher, Larry. Larry also matches Buster. I was very excited to see this three way match and I wrote to Larry asking if he had a Moore line. Yes, he did, two in fact. The Levi Moore line out of Kentucky and an Alexander Moore line out of Stokes County, NC, after they wandered down from Berks Co., PA. sometime before 1803.
Groan. Two Moores – I can’t even manage to sort one out, how will I ever sort two?
Then Larry told me that he had 4 of his cousins tested too. Bless you Larry.
And better yet, one of Larry’s Moore lines is on his mother’s side and one on his father’s. Even better yet. Things are improving.
Now I’m really excited, right up until I discover that my cousin Buster matches two of Larry’s 3 cousins on his mother’s side and my Moore cousin from Halifax County, Virginia, matches the cousin on Larry’s father’s side.
How could I be THIS unlucky???
So I started out utilizing the ICW and Matrix tools at Family Tree DNA. Because these people all matched Larry on overlapping segments on the chromosome browser, my first thought was maybe that these two Moore lines were really one and the same. But then I pushed the ICW matches through to the Family Finder Matrix, and no, Larry’s paternal cousin does not match any of the three maternal cousins, who all match each other. So the two Moore families are not one and the same.
Crumb. Thank Heavens though for the Matrix which provides proof positive of whether your matches match each other. Remember, you have two sides to each chromosome and you will have matches to both sides. Without the Matrix tool, you have no way of knowing which of your matches are from the same side of your chromosome, meaning Mom’s side or Dad’s side.
Just about this time, as I was beginning to construct matrixes of who matches whom in the ICW compares between all of the ICW match permutations, I received a note from Don that he wanted beta testers for his new ADSA application. I immediately knew what I was going to test!
I started with my cousin Buster’s kit. Buster is one generation upstream from me, so one generation closer to the Moore ancestors.
On Larry’s maternal line, descended from the Levi Moore (Ky) line, he tested three cousins. Buster had the following match results with Larry and his maternal line cousins.
- Larry – match
- Janice – no match
- Ronald – match
- B.J. – match
I have redacted the e-mails and surnames below, but want to draw your attention to the individuals with the red arrows, as noted above.
On the graphic below, I’m showing only the right side, so you can see the matching ICW (in common with) block patterns. Larry is last, I’m second from last and Larry’s two cousins are the first and second red arrows. We are all matching to my cousin, Buster.
You can see that all of these people match Buster. Larry has blocks that are pink, red, fuchsia, gold, navy blue and lime green. All of the group above, except me and two other people, one of which is my known cousin on another line, match Larry on these blocks, or at least most of these blocks. I, however, match none of this group on none of these blocks, nor do my other known cousins who also descend through this same Moore line. This means that this group matches Buster through Buster’s mother’s line, not through the Estes line, which means that this Moore line is not the James Moore line of Halifax County. So the Levi Moore group of Kentucky is not descended from the James Moore group of Prince Edward and Halifax County.
Of course, I’m disappointed, but eliminating possibilities is just as important as confirming them. I keep telling myself that anyway.
The male Moore descendant in Halifax Co., proven via Y line testing, does match with Chloa, Larry’s paternal cousin, and with Larry as well, as shown below. Let’s see if we can discern any other people who match in a cluster, which would give us other people to contact about their Moore lines. Keep in mind that we don’t know that the DNA in common here is from the Moore line. It could come from another common line. That is part of what we’d like to prove.
Let’s take a closer look at what this is telling us.
First, there’s a much smaller group, and this is the only chromosome where Chloa matches our Moore cousin.
So let’s look at each line. The first person, John, doesn’t match anyone else, so he’s not in this group.
Larry and his cousin, Chloa are second and third from the bottom, and they form the match group. You can see that they match exactly except Chloa has one brighter green segment that matches our Moore cousin in a location with no other matches. However, the match group of navy blue, periwinkle, lime green and burgundy form a distinctive pattern. In addition to Chloa and Larry, Virginia, and Arlina share the same segments, plus Arlina had a pink segment that Larry and Chloa don’t have. Donald may be a cousin too, but we don’t know if Donald would also match the rest of the group. Linda might match Donald, but doesn’t look like she matches the group, but she could. At this point, we can drop back to Family Tree DNA and the matrix and take a look to see if these folks match each other in the way we’d expect based on the ADSA tool.
Just like we expected, John doesn’t match anyone. As expected, Larry, Chloa, Arlina, and Virginia all matched each other. As it turns out, Linda does not match the rest of the group, but she does match Donald, who does match Arlina. Therefore, our focus needs to be on contacting Arlina, Donald and Virginia and asking them about their Moore lines, or the surnames of known Moore wives, such as Rice in my James Moore line or wives surnames in Larry’s Moore line. Just on the basis of possibility, I would also contact Linda and ask, but she is the long shot. However, like the lottery, you can’t win if you don’t play, so just send that one extra e-mail. You never know. Life is made up of stories about serendipity and opportunities almost missed.
If Larry’s Moore line is the same as our Moore cousin’s line, genetically, maybe we can make headway by tracking Larry’s line. Larry was kind enough to provide me with a website, and his Moore line begins with daughter Sarah. Her father is Alexander Moore born in 1730 who married Elizabeth Wright. His father was Alexander born in 1710 and who lived in Bucks Co., PA. The younger Alexander died in Stokes Co., NC in 1803.
Our next step is to see if this Alexander Moore line has been Y DNA tested. Checking back at the Moore Worldwide project, this family line is not showing, but I’ve dropped a note to the administrators, just the same. Unfortunately, not everyone enters their most distant ancestor information which means that information is blank on the project website.
If this Alexander Moore line has been Y tested, then we already know they don’t match our group paternally. The connection, in that case, if this genetic connection is a Moore line, could be due to a daughter birth. If this Moore line has not been Y tested, then it means that I’ll be trying to track down a Moore descendant of one of these Alexander Moores to do the DNA test. It would be wonderful to finally make some headway on the James Moore family. We’ve been brick walled for such a long time.
If you descend from either of these Moore family lines, the James Moore (c 1720-c 1798) and Mary Rice line, or the Alexander Moore and Elizabeth Wright or Elizabeth Robinson line, please consider taking the Family Finder autosomal DNA test at Family Tree DNA. If you know of a male Moore who descends from the Alexander Moore line, let’s see if he would be willing to Y DNA test.
There is a great deal of power in the combined results of descendants, as you can clearly see, thanks to Don Worth and his new Autosomal DNA Segment Analyzer tool.
Give it a test run at: http://www.DNAgedcom.com/adsa
Don wrote documentation and instructions, found here. Please read them before downloading your files.
And Don, a big, hearty thank you for this new way to “see” our ancestors! Thank you to Rob Warthen too for hosting this wonderful new tool!
______________________________________________________________
Disclosure
I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.
Thank you so much.
DNA Purchases and Free Transfers
- Family Tree DNA
- MyHeritage DNA only
- MyHeritage DNA plus Health
- MyHeritage FREE DNA file upload
- AncestryDNA
- 23andMe Ancestry
- 23andMe Ancestry Plus Health
- LivingDNA
Genealogy Services
Genealogy Research
- Legacy Tree Genealogists for genealogy research
Looks like fun. I have playing with the generation calculator on Gedmatch. It looks like I am only one generation away from making a connection. I am sure this will help too.
Roberta I have to say it….I love you girl….once again, you have swooped in and saved me a ton of work! Just last night, as part of my New Years Resolution, I was working on building a master spreadsheet with all of my matches at FTDNA, in Chromosome, then Segment order, then I was going to move on to the “ICW”; then surnames, then emailing. And then I was going to the gedmatch matches….oh my! I literally swooned when I saw this!
So now, my New Years Resolution, is to use this tool immediately, create a fan chart with all my X chromosomes filled in, write a blog about the importance of sharing information, and send the whole kit and caboodle to each one of my matches, gedcom uploaded or not! And, oh yes, in my blog I will be highly recommending that they all read your blog.
And not really related to this topic, but in a way it is…hahaha….I had one of those aha moments this week. I was contacted my an elderly lady that used to be on my match list, but no longer is, asking if I had any Futrell’s in my Bridges/Cohoon lines. I didn’t. Luckily, she uploaded to gedmatch, and that site predicts 7.2 generations to our common ancestor. A needle in a haystack? At first, I thought so, but, further investigation and questioning revealed that she was working with a geneticist, she sent me the fan chart that was prepared for her with the Futrell they think she descends from and, low and behold, there was Priscilla Flowers! My 7th great grandmother!. Priscilla Flowers married John Calvin Cohoon. Their daughter, Charity, married Drury Bridger/Bridges. My match descends through another daughter of this couple who married a Futrell. So our connection wasn’t Bridges/Cohoon as she at first thought, but was Flowers/Cohoon.
So! Sometimes, you get lucky and know to ask the right question, but with a fan chart, it kind of hits you in the face.
Wow, I think I’m excited!
Carla Wallain Davis
________________________________
Thank you for posting this! It IS exciting!!
Dear Kay,
Something to do with your Autosomal DNA results from FTDNA. I’m interested in using the Analyzer with my Karr and Merville results.
With best regards,
Leslie
This is so crazy! I sat down yesterday to start chromosome mapping for my family and concluded I needed exactly this tool! (Imagine how “visually oriented” I am as a graphic designer.) I was quizzing geeks of my acquaintance on how it might be done —and somebody has already done it! Hallelujah!
Great (visual) minds think alike, eh? 🙂
Kudos. Beautiful. Thank you.
This tool is exactly the kind of that I have had in mind when I’ve complained that the DNA testing companies have done a poor job in making their tools as useful and intuitive as they need to be. As usual, the amateur wizards are a step ahead.
Oops. I’m gonna have to stop trying to comment with my kinds climbing on my shoulders and bouncing in my lap, but I suppose the point is clear enough.
So 23andme is next? And maybe Ancestry, if they ever unleash the matching segment data?
Yes, Rob (DNAgedcom) plans to come up with a merged data model that would support 23andMe and Ancestry etc. I will modify ADSA to access that when its available. However, I don’t think 23andMe has ICW data available does it?
Don,
They don’t call it ICW data, but yet, you can download each of your matches matches to your other matches, but it has to be done one by one at this time using Rob’s tools. I bet he can come up with a tool to run through them all in one fell swoop though.
Oh – now I remember you mentioned that to me earlier. Well, when it comes to the data, Rob is my “main man”. I will defer to him (and pass the buck). 🙂
You can use my files. They are all on Rob’s site:)
Great. Thanks again for the fantastic work. Yes, 23andMe has data essentially equivalent to ICW.
This analyzer is already an embarrassment of riches, but if there was a way to generate lists of surnames SHARED by overlapping cousins, that would be awesome. The names are embedded in the spreadsheet, of course, but comparing the lists of names manually is a bit tricky. Maybe there’s already some easy way to do this?
Yes, I agree; I wrote to FTDNA about doing just that, but I may not live long enough to see them do it.
I’m planning to match to the person tested’s GED file at some point and bold the names that match (as FTDNA does) but it doesn’t sound like that’s precisely what you are after? The problem with doing things on the basis of overlapping segments is that I’m not sure how to group the segments. How much should they overlap to be “overlapping”, do they all have to ICW to one another? Ultimately I think that’s kind of a judgement call for the person using the tool probably.
Just a little warning. I did this tool partly because I thought there were a lot of people out there who didn’t know (or want to know) how to use spreadsheets. Ironically, a problem we discovered today has as a workaround using a spreadsheet. Sigh…
For a few people (4 that I talked to today) the ICW file is too large and the server cuts them off. They get either a 500 server error message or a blank screen. I have seen this where the ICW file is over 4 or 5 MB. There was one today that was over 10MB. If you have a lot of your close relatives tested, it may happen to you. It turns out that there is a lot of data in the ICW file that is simply a copy of what is already in the MATCHES file and can be safely deleted. (I’m not sure why it was there in the first place)
I recommend for anyone with spreadsheet savvy – open the ICW file, delete all the columns EXCEPT the first two (the ones with the ICW names in them), and save it back as a .csv file (Mac users select WINDOWS CSV from the drop down list). You can add something to the end of the file name if you want to preserve the original file. Then, when you use the ADSA tool, use the new ICW file instead of the original one. I had a 6.5 MB ICW file in one of my kits – it went down to 500KB. Even if the tool is working for you, doing this should greatly speed things up!
I have discussed this with Rob at DNAgedcom and he plans to build this into his download tool sometime in the next few days.
Don
I’m not sure I understand. Are we supposed to click on the FTDNA button at the top of the DNAgedcom website for the download? That’s what I’ve been doing and my file must be too large because it stops every time around the 67% completion mark. The screen either goes blank or the down load process simply starts over again on its own.
I don’t really understand the directions above using Excel. I have downloaded all my matches to Excel but don’t understand about downloading ICW . Would that not require on checking every matches ICW and down loading them to Excel? Also, where is the option on the DNAgedcom site to download this new format instead of downloading from the FTDNA button.
As you can see I’m having a difficult time understanding.
Teresa – I was explaining that when you are UPLOADING the files to ADSA you need to first delete away all those extra columns to keep the ICW file size down. When you are DOWNLOADING the files from DNAgedcom, you want to get all three files in a zip file – which will include an ICW file that was obtained by DNAgedcom from FTDNA. Sometimes this process doesn’t work quite right – maybe it didn’t for you. In that case, I believe the people at DNAgedcom are recommending that you logon to their site and then click on the MEMBERS button at the top right of the screen and select VIEW FILES. Click on the kit number you were attempting to download and its files should be listed there. You can download each file (unzipped) one-by-one by double-clicking on the filename. That might work where the all-in-one download didn’t. If you run into trouble, contact DNAgedcom for help through their web site. They are very helpful!
Don
Don, thank you, so much for the explanation. It was easy after finding the files under the ‘Member’ tab. Everything seems to be working and what a tool!! I can already see that it’s going to help me.
Terry
This is so exciting. However, the download has stopped completely and vanished from the page four times. I’ll try again tomorrow. 🙂
A truly amazing and wonderful tool, thank you so much to all involved in the creation!
I’m a bit gobsmacked, my mother is adopted and I know this tool is going to help break the matches into maternal/paternal On her Chromosome 9 she has 90 matches and Chromosome 10 there are 35, and for me, Chrom 9 = 36 and Chrom 10 = 35, is that 90 normal? Compared to the other cousins I’ve looked at its excessively high, how should I interpret that?
You know, we do talk about “sticky segments” that seems to inexplicably get passed in tact for many generations. Maybe that?
Thanks Roberta, I will do some research.
Do the matching color-coded ICW blocks imply that the ICW match is for that match on that chromosome only or on any chromosome? As its currently set up, the ICW isn’t that useful if multiple people match you and/or each other on multiple segments. I’d say the same thing about the FTDNA matrix tool. Is there a way to eliminate this uncertainty using these tools?
It’s on the exact same part of the same chromosome. You can eliminate the ambiguity between both the ICW and the matrix together.
So, there should be a common ancestor?
If you have an ICW match, with several people on the same segment of DNA, you know that they have to match you on one side or the other – Mom or Dad. If they also match each other, then that indicates that they share a common ancestor. This is what you utilize the FTDNA Matrix for. If they match ICW but don’t match each other, either the segment is too small to match or they don’t share a common ancestor between them.
Expanding on what Roberta said, my understanding is that the ICW is indicating simply “this one of my matches also matches this other one of my matches – some way, somehow”. There is no implication about where the match occurred. Hence the fact that you can have two matches who ICW each other, but you don’t see their segments on the same chromosome. So the ICW just tells you that there is a match – but it is related to the people, not (necessarily) the segment you are looking at. Of course if you match someone on one 15 cM segment and someone else on a segment that overlaps it, and they are ICW to each other, that’s a pretty big clue that that segment had something to do with their having been declared as matching each other and you (although they could also share more cMs somewhere else that you can’t see because they share them with each other, but not with you).
ADSA Updates
Rob and I have been busy the past day or so. The main thing that’s been driving us has been to find a way to make ADSA available to those with Ashkenazi ancestry. As you may know, their ICW files, in particular, are tremendously large (300MB or more) and ADSA was choking on them and timing out or giving 500 Internal Server Errors. In the course of working something out for the Ashkenazi community, we have solved some other problems and added enhancements that will benefit everyone. A nice win win.
First, when you go into the initial ADSA screen now you have two options. If you are not logged on to DNAgedcom you will see the prompts for the three files on your computer’s hard drive as before. However, if you logon to DNAgedcom and then navigate to ADSA, these prompts will be replaced by a prompt for a kit number. If you are logged on, ADSA will go find the files on the DNAgedcom server, eliminating the upload step of the process and speeding things up by a factor of 10. This was needed to read huge files, but it makes things faster for everyone!
Rob has overhauled the ICW file download to eliminate redundant data that FTDNA had included in that file (duplicated match data was on every row). If you re-download your zip files now you will find that the ICW is less that 10% the size it once was. Doing this made supporting Ashkenazi kits possible.
I have also added some new input fields on the form for ADSA. If you specify a single chromosome to map now, you can give a starting and ending base pair location. ADSA will then only graph segments that overlap that range. This can be useful if you want to lens down to a small segment and lower the minimum cMs to 1 cM, for example.
I have also added a most distant cousin limiter. If you enter 4, for example, ADSA will ignore any segments and matches further out than predicted 4th cousin. This is what really makes it possible to read an Ashkenazi kit. We have tested it on 4 or 5 kits. If you have Ashkenazi ancestry, download your kit(s) again now, check VIEW FILES in DNAgedcom to make sure they are there (and note the size of your ICW file – should be smaller than before), then go to ADSA while logged on and choose 3rd or 4th cousin and enter you kit number. I will have my fingers crossed that it will work for you! If you run into any problems, contact me or support@dnagedcom.com. But I have high hopes we’ve whipped the Ahskenazi challenge! Yay!
I have updated the documentation accordingly, so you can read more about these new features there. There is also a new section on troubleshooting based on our collective experience over the past few days since ASDA was released.
Let mw know what you think!
Don
http://www.dnagedcom.com/adsa
I’m not Ashkenazi, but bless you nonetheless.
I have fallen IN LOVE with this tool. It seems to be a combination of a Chromosome Browser and Matrix Matcher on steroids. Now if someone could just combine it with a by-chromosome Admixture painter, wouldn’t we be in DNA-heaven?
Tell me how to do that and I will see what I can do. *wink*
From one software developer to another – I challenge you!
I don’t know if this is the intended use for this tool, but here is the experiment I am trying: I have a group of 6 people who match me on 6 segments on the same chromosome. Several of these people included a long list of surnames in their FTDNA file and not a single surname matches my paper genealogy research. I have one troublesome genealogy dead end on my mother’s side, and the geographic locations listed by the other people suggest this is where the match is. Either that or I’ve got a non-paternal event somewhere that I don’t know about. So, my experiment is e-mailing all six of these people and explaining all of the above and waiting to see if anyone of them is interested in further research.
Actually, I think this is exactly how this tool was intended to be used:) To solve mysteries!
I agree with Roberta – it never hurts to email one of your matches even if all you have is a hunch. That’s why I think the ICWs of one of your matches who may match on different segments and chromosomes entirely could share the same ancestor (you can see a list of their names by hovering your cursor over one of the colored horizontal segments). In my humble opinion (and it is humble!) ICWing a group without everyone matching you on the same segment gives you a very big hint that they may all share the same ancestor. The reverse is not true – you will want to rule out anyone who doesn’t match each other with the ICW. Just because they all match you and have segments on the same chromosome, if they are separate segments and they don’t match each other (some way, some how), then they probably all match a different one of your many ancestors. They could even be on the opposite side of your family, even if they DO match on the same segment.
Don
Pingback: What If You Die? | DNAeXplained – Genetic Genealogy
I have what is probably a really dumb problem. I have done the Genographic project, Family Tree mtDNA hvr1 and hvr2 and now Geno 2. I got my results back a couple of days ago and downloaded them to Family Tree. What I wanted most to see was the autosomal bits like you’ve been writing about and showing. But nothing is showing. Because of a back mutation, I have no matches and never will. The back mutation is in the coding region of the mtDNA so my haplogroup cannot really even be designated. Do you have to have matches for this to show up? Do i have to buy yet another test from Family Tree? What am I doing wrong? If you have to have matches, I’m screwed.
Thanks
You have several questions here.
1. mtDNA – the geno project only provides a haplogroup. Family Tree DNA does the individual tests to match people. Have you taken those tests. You won’t match anyone with just the Geno haplogroup.
2. The autosomal results from National Geographic are for deep ancestry and are not the same results that you get from the Family Finder test at Family Tree DNA. The Family Finder test is the one that provides the files to use with the ADSA.
I am not really trying to find matches; I understand that there will never be any (except my daughters, presumably). And everyone is really mealy-mouthed about my haplogroup. Well you’re kinda, sorta mumble we need to do more research. Fine. But I was really hoping to see some, I don’t know, numbers, hard facts, something. Deep ancestry is fine, I don’t care. The ISOGG website implies that you can download the raw data from Geno 2 and play with the ethnicity markers. Like, what are those markers, how many are there for different groups? Is this possible?
You can download your data from Geno 2.0, but we really don’t utilize it for other things. I don’t think Gedmatch accepts it. Most people utilize Family Tree DNA for the raw data files.
I love this tool. I’m a visual person – spreadsheets make me whimper. I was hoping to find that we’re a match on the Moore line – I have them in Gaston/Lincoln counties – but no such luck.
I want to love this tool, too! I have tool envy.
I have about 17 people (and counting) who match along a segment of chromosome 7. Although they all match me, subsets of the group match each other. (I can subdivide them into groups A, B, C, and D where A matches A, B matches B, C matches C, etc. Some of B matches A and C, some of C matches A and B; A, B, and D on the segment, etc. and some match each other (but not me) on more than one segment).
I would love to have a graphical interface to look at these matches along my chromosome. How soon to see this tool available for the 23andmers?
I would love the same thing!
Rob Warthen and I are working on it but I’m guessing it will take at least. A couple of months. It’s coming with a complete rewrite of ADSA to make it more flexible, so we will have to do quite a bit of testing to make sure it’s working well.
Don
I was doing this using the gedmatch chromosome matching tool. Then this tool stopped functioning. It is very handy and hopefully it will be available again.
Hi, what a great tool. Question – I have two people that match on an overlapping segment on chr. 4 with a 13 cM overlap. But they are NOT indicated ICW to one another. Does that mean that they each match different chr 4’s from me, but are completely not in common with each other at that location? Because otherwise I would think a match as long as 13 cM would indicate identical by descent. thanks for your input.
Hi Michelle. Do you mean they both overlap you with a 13cM segment? If they overlap each other with that segment, then they would both be ICW with you if all 3 of you match on that segment. I think I’m misunderstanding something in your question.
Hi Roberta,
Hope you can see this makes it thru to you. This is my kit. The pink and yellow here overlap by 13.35 cM but they are NOT ICW. I assume this means there is a 13.35 cM stretch where Yellow matches me AND Pink matches me. But since Pink and Yellow are not ICW each other as indicated on the left, am I to assume that 1) a 13.35 cM match is not enough to be considered cousins or 2) that they both match me (each matching me in identical by half fashion, but mis-matching to each other, or said in another way, I match Pink on one of my chromosome 4âs but match Yellow on the other.
Thanks again for any help⦠the explanation might be of general use to the audience, too. Sending to you here the picture comes thru.
Michelle
The picture did not come through, but the answer is that it could be either one. FTDNA does require a total match of 20cM to be considered a match – so maybe that is it. I’d suggest asking if all the people involved will download their results to GedMatch and see what each person looks like there compared to the other one on that segment.
Roberta
Also FTDNA requires that two people share 20 cMs of DNA overall for a match (ICW). Maybe that 14 cM is all they have.
thanks Roberta, that answers the question… now to convince them both to post their gedcoms….
From a family tree I found on omii.org (Ohio, Michigan, Indiana, Illinois)years ago I found my ancestor Henry Sauder b 1802 as the brother to Joseph Sauder b1805. Henry moved to Indiana while Joseph stayed in Ohio. Yesterday I matched a lady as a 4th cousin 1xR and she is descended from Joseph. So the segment analyzer works. We matched at 23.91.
Hi All — I got some new results in and went to use the JWorks and autosomal dna segment analyzer, but now DNAgedcom is saying that the site no longer works with Family Tree DNA. Does anyone know what’s up with that? Is there a way to download the ICW file directly from Family Tree DNA?
Replies appreciated!
I think DNAgedcom has an email address on their site for customer support. I might be confusing that with another site, though.
The support address is support@DNAgedcom.com – but they will tell you more or less what I said below. 🙂
I have been in contact with the site. Apparently FTDNA tweaked their interface, so now DNAgedcom has to adjust their programming. They’re hoping it won’t take long. You should still be able to use the ADSA with files you have downloaded directly to your computer, though.
Thanks Don and Valerie!
DNAgedcom has had some issues of late w/re to interdependencies with FamilyTreeDNA as it downloads match data. There have been some problems where the downloads can overload FTDNA’s servers and Rob Warthen (who developed the downloader) has been working with FTDNA to manage the process to reduce the impact. Apparently a problem cropped up while Rob was out of the country on vacation and FTDNA shut down the connection to DNAgedcom. Rob is back now and is working on a fix. Hopefully the service will be restored soon – however Rob doesn’t have a specific ETA yet.
Is there something unusual about chromosome 12? On that chromosome I have a large number of people (89) with small matching segments, all beginning at 20875620.
39 end at 21525661, with a length of 2.46 cM (500 SNPs)
21 end at 21882887, with a length of 3.07 cM (600 SNPs)
21 end at 22128857, with a length of 3.71 cM (700 SNPs)
7 end at 22538447, with a length of 4.11 cM (800 SNPs)
1 end at 23169216, with a length of 5.17 cM (1000 SNPs)
Does this mean that all of these people have a common ancestor or, considering the small sizes involved, could these be IBS matches?
I’m mainly concerned with one of these 89 people who doesn’t appear as a match to me or my mother but does appear as a match to my mum’s half-sister. They share a father and we are trying to determine who her mother was.
It could be IBS, which could just be an ancestor further back in time that you’re not going to be able to find. It could also be IBD, or some combination of both. I know this isn’t very helpful.
I have found that a lot of times you can see these large groups of overlapping segments – especially in the under 4 cM range. Since the vast majority of segments that small are IBS, and even if the segment is IBD short segments usually denote a very distant most recent common ancestor (MRCA), I usually don’t bother with them. I suppose if they overlap a longer set of segments that are most likely IBD (10 cM or more) which triangulate each other, and there are matches between the owners of the short segments and those in the Triangulated Group, you might give them more credence – especially if you can find the MRCA of that TG and he/she is the ancestor you were expecting them to be connected to. I have read recently that as many as 20% of the very short segments might be IBD. The hard part is figuring out which are IBD and you can’t use triangulation to ensure you’re looking at a match on the right chromosome of the pair with very small segments like this.
Thank you for your replies. These short segments seem to be the only place where I match person of interest, so he may be worthy of investigation as a maternal match to mum’s half-sister. This is very helpful.
I only have one match in this area. The specific snp is small also. But I also have 16.1 match with him on the 12th chromosome. He is primarily Scots from Australia.
Sarah Moore, I’m your cousin through all those Acadian names !
Pingback: 2014 Top Genetic Genealogy Happenings – A Baker’s Dozen +1 | DNAeXplained – Genetic Genealogy
Pingback: DNA Testing Strategy for Adoptees and People with Uncertain Parentage | DNAeXplained – Genetic Genealogy
Pingback: DNA Data Organization, Tools and Who’s on First? | DNAeXplained – Genetic Genealogy
Pingback: DNAeXplain Archives – Intermediate DNA Articles | DNAeXplained – Genetic Genealogy
Dear Roberta!
I have tried the ADSA tool. I get quite a lot overlapping segments on all chromosomes, but no ICW. One typical example: http://joxi.ru/a2XaLd1fy397zA
Based on the comments above, there could be 2 reasons:
1) the segments are not long enough to be considered cousins? But I selected only segments larger than 20 cM. Is that still not enough.
2) If there were only 2 people matching, we could say that one of them matches the segment on father’s side, and the other matches the segment on mother’s side. But now we have 8 people. So, clearly, more than one must be from the same side.
One more detail: the first 2 matches are respectively mother and son of the person being analysed. They are grandmother and grandson to each other, and still they are not indicated as matches. How could that be?
Thanks,
Mantvydas
Hi Mantvydas,
It looks like the kit didn’t entirely download and you didn’t get any ICWs. It could be a timeout problem. Send me an email at worth@ucla.edu with your DNAgedcom username and I will look into it for you.
Don Worth
DNAgedcom (ADSA author)
Pingback: Concepts – The Faces of Endogamy | DNAeXplained – Genetic Genealogy