World Families Forums - What is considered “significantly off-modal” for L21

Welcome, Guest. Please login or register.
December 19, 2014, 04:38:20 AM
Home Help Search Login Register

+  World Families Forums
|-+  General Forums - Note: You must Be Logged In to post. Anyone can browse.
| |-+  R1b General (Moderator: rms2)
| | |-+  What is considered “significantly off-modal” for L21
« previous next »
Pages: [1] Go Down Print
Author Topic: What is considered “significantly off-modal” for L21  (Read 1531 times)
admixman
Member
**
Offline Offline

Posts: 8


« on: July 06, 2011, 11:06:53 PM »

I recently got my deep clade results from FTDNA, and I am R-L21. However, I don’t seem to match the L21 modal very well. For 37 markers, I have a genetic distance of 18 off the L21 modal; for 67 markers, the genetic distance is 20. In fact, I have no matches in the FTDNA R-L21 Project, even for 12 markers (see kit #188436). I have several Y STR markers that are unusual for R1b - DYS393=12, DYS385b=16, DYS458=16, DYS437=16, DYS406S1=12, & DYS617=13.

I have not been able to trace my paternal line back within Europe, but the most likely ancestries are German, Italian, Turkic, or Russian. I have done autosomal DNA tests with DNA Tribes, DNA Consultants, and GTL Ancestral Origins. All of them point to strong Italian, Turkic, and Russian influences, which are unaccounted for in my known genealogy. 

Questions:
1. What is considered significantly off-modal for a haplogroup?
2. Is the DYS393=12 necessarily a back mutation for someone who is L21, or could it indicate an Eastern European origin?
3. Is there any particular significance to the other unusual markers?
4. Any ideas on the likely origin of my paternal line?
Logged
rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #1 on: July 07, 2011, 09:16:03 AM »

I recently got my deep clade results from FTDNA, and I am R-L21. However, I don’t seem to match the L21 modal very well. For 37 markers, I have a genetic distance of 18 off the L21 modal; for 67 markers, the genetic distance is 20. In fact, I have no matches in the FTDNA R-L21 Project, even for 12 markers (see kit #188436). I have several Y STR markers that are unusual for R1b - DYS393=12, DYS385b=16, DYS458=16, DYS437=16, DYS406S1=12, & DYS617=13.

I have not been able to trace my paternal line back within Europe, but the most likely ancestries are German, Italian, Turkic, or Russian. I have done autosomal DNA tests with DNA Tribes, DNA Consultants, and GTL Ancestral Origins. All of them point to strong Italian, Turkic, and Russian influences, which are unaccounted for in my known genealogy.  

Questions:
1. What is considered significantly off-modal for a haplogroup?
2. Is the DYS393=12 necessarily a back mutation for someone who is L21, or could it indicate an Eastern European origin?
3. Is there any particular significance to the other unusual markers?
4. Any ideas on the likely origin of my paternal line?


1. The L21 modal haplotype, last I checked, was pretty much the "Super Western Atlantic Modal Haplotype", meaning an extended version of FTDNA's WAMH, which was an extended version of the old six-marker Atlantic Modal Haplotype, which, of course, is the modal haplotype for Western Europe.

A modal haplotype is just a collection of the most frequently occurring marker values in a set or collection of haplotypes. The longer the modal haplotype, the less likely anyone is to match it exactly. In fact, beyond the first six or twelve markers, I would be really surprised if anyone matches the L21 modal haplotype exactly. Everybody has some off-modal markers. Being off it by 20 at 67 markers is not at all unusual.

2. Yes, 393=12 is a back mutation for those of us who are L21+, since L21 is L11+. 393=12 as a modal value is a characteristic of men who are L11- and marks the progress of R-L51 from the east or southeast to the west and northwest across Europe. At some point, probably in eastern or central Europe, the transition to L11+ occurred and, with it, the move to 393=13.

c. Regarding your 406S1=12 and 617=13, that means you have the "11-13" combo, but your 406S1 value has moved up one to 12. That combination is widespread across Western Europe and seems to be universally L513+ (L513 is downstream of L21). L513 has a couple of subclades which are characterized by the SNPs L193 and P66. If you have not yet been tested for those SNPs, you should consider it.

Otherwise, it will take some time and research to tell what the significance of your off-modal markers is. Go to Ysearch and use those markers, plus a couple of common R1b markers, to create an 8-marker haplotype (the minimum Ysearch will accept). Run that short haplotype for matches and see what you get.  Exclude anyone who is outside R1b (like R1a, which will probably provide you with some haplotype "noise").

4. For your paternal line, since you are L21+, the answer is mostly likely Western Europe, probably among the Bronze Age Celts, but that is not absolutely 100% certain. As for your autosomal results, forget them - well, don't forget them - but they are of no consequence in trying to ferret out information on your straight y-dna line.

First off, half of your ancestors are female, and they contributed half of your dna, including half of your autosomal dna. Secondly, your paternal or y-dna line is only one, single line. Even though your father contributed half of your dna, he himself got his own dna from a combination of a multitude of ancestors, again, half of them female. The y chromosome is completely separate from autosomal dna.

We have some African American guys in the R-L21 Plus Project. I dare say that if you saw them on the street, you would never guess they were R-L21, but they are. They look African because the cumulative effect of their autosomal dna gives them an African physical appearance and probably, in terms of dna, they are mostly African. But that one, single, paternal, y-dna line is European.

Since autosomal dna recombines and gets passed on willy-nilly (look, I know it follows the dominant-recessive rules), there's no telling which of your multitude of ancestors contributed which bit of autosomal dna. Autosomally, you could be a Russian or a Finn or whatever. But in your y-dna you are a Western European.

« Last Edit: July 07, 2011, 10:11:30 AM by rms2 » Logged

Mike Walsh
Guru
*****
Offline Offline

Posts: 2964


WWW
« Reply #2 on: July 07, 2011, 03:29:28 PM »

.... In fact, I have no matches in the FTDNA R-L21 Project, even for 12 markers (see kit #188436). I have several Y STR markers that are unusual for R1b - DYS393=12, DYS385b=16, DYS458=16, DYS437=16, DYS406S1=12, & DYS617=13.
... . Is there any particular significance to the other unusual markers?
Admixman,
Do you have a Ysearch ID?  I recommend creating one. You can do it from your FTDNA Y DNA Matches screen.

As RMS indicated, you could be classified as in the "11-13 Combo" group because of your unusual values at two fairly slow markers, DYS617=13 and DYS406s1>=11.

I have your haplotype loaded in a spreadsheet of L21+ people and you come out closest to the following, but I wouldn't say you have a genealogical timeframe relationship with anyone of the below.... probably a "clan" or "tribal" type relationship of 1000 years old or more.  Still, this could help you find your paternal lineage's geographic origin.

You apparently have a lot in common with "11-13" Group K, which is mostly the Kingston surnamed people.  Two people in 11-13 Group K tested positive for a new SNP named L513  so you might consider that test.  That would help determine your most recent branch on the Y DNA family tree. FTDNA calls it a haplotree.

Please join our project at http://www.familytreedna.com/public/R-L21-1113Combo/


f188436___ zzzUnkName_______________ R-L21____________ 1113___________ ___ zzzUnkOrigin
f48440____ Kenny____________________ zzPredicted______ 1113-?K________ Y34QR___ Ireland, Connacht, Co. Galway, Kilconnell (or Co. Mayo, Ballyglass)
y9JKHQ____ Wiley____________________ zzPredicted______ 1113-?K________ 9JKHQ___ zzzUnkOrigin
f35500____ McAlister________________ zzPredicted______ 1113-K_________ ___ zzzUnkOrigin
f76048____ McAllister_______________ R-L21____________ 1113-K_________ CAAGC___ Ireland, Ulster, Co. Antrim
f147680___ Diamond _________________ R-L21____________ 1113-O_________ T5X8E___ Ireland, Connacht, Co. Sligo, Skreen
f57445____ Gettings_________________ zzPredicted______ 1113-K_________ ___ Ireland
f170539___ Kingston_________________ zzPredicted______ 1113-K_________ NVAXR___ England, East, Bedfordshire, Cople
f114954___ Kingston_________________ zzPredicted______ 1113-K_________ ___ Ireland
f173705___ Kingston_________________ zzPredicted______ 1113-K_________ ___ England, East Midlands, Northamptonshire, Silverstone
f106714___ Kingston_________________ zzPredicted______ 1113-K_________ ___ Ireland, Munster, Co. Cork, Drimoleague
f179680___ Kingston_________________ zzPredicted______ 1113-K_________ ___ England, East Midlands, Northamptonshire, Paulerspury
f107275___ Kingston_________________ zzPredicted______ 1113-K_________ ___ Ireland, Munster, Co. Cork, Drimoleague
f186680___ Kingston_________________ zzPredicted______ 1113-K_________ ___ Ireland, Munster, Co. Cork, Timoleague


You can't order L513 from the normal SNP haplotree page. You have go to in through this path.

Login to your FTDNA account to get to your MY ACCOUNT page.
Under MY ACCOUNT, click on ORDER TESTS & UPGRADES.
From the ORDER ADDITIONAL DNA TESTs page scroll down and click on the orange button for ORDER ADVANCED TESTS* button.
On the RETURNING CUSTOMER – UPGRADING KIT page, under SELECT A PRODUCT pick SNP in the TEST TYPE drop down arrow menu.
In the MARKER box type in the name* of the SNP and click on FIND.
A new row on the screen should pop up that looks like "SNP L513 $29.00 Add."
Just click on ADD and continue and pay.

It seems strange, but you don't order from the ADVANCED SNP TESTS orange button, but the one below that says ORDER ADVANCED TESTS.

Please join the 11-13 project at
« Last Edit: July 07, 2011, 03:30:57 PM by Mikewww » Logged

R1b-L21>L513(DF1)>S6365>L705.2(&CTS11744,CTS6621)
admixman
Member
**
Offline Offline

Posts: 8


« Reply #3 on: July 08, 2011, 02:22:19 AM »

RMS,
Thank you for your rapid response to my questions and for your suggestions.

In order to better understand the significance of genetic distance from the modal, I have been studying Mike Walsch’s latest Excel spreadsheet for L21+ tested haplotypes and Alex Williamson’s Phylogenetic Tree for R-L21, at:

     www.littlescottishcluster.com/RL21/Files/

Both are very impressive pieces of work. I must admit that the L21+ spreadsheet brought my Apple MacBook to it’s knees!

Based on the latest version of the L21 modal, my 67-marker haplotype has a genetic distance of 21 off the modal. Searching through the database there are 102 members out of a total of 2715, who have a genetic distance greater than 20 from the modal, or about 4%. There is 1 at 26, 1 at 27, and 1 at 29.

Although deciding what is “significantly off-modal” may be subjective, I am comfortable with the greater than 20 and 4% criteria.

Alex Williamson calculates an age estimate for R-L21 of 3420±484 years, based on a mutation rate of 0.002 for all loci. I have not attempted the math, but I suspect that those of us who are significantly off-modal at slow-mutating loci, could bump up against that age estimate, if we use loci-specific mutation rates. Four of my five unusual alleles have mutation rates less than 0.002 (DYS393=0.0007, DYS437=0.0009, DYS406=0.0015, DYS617=0.00042).

Following your suggestion, I tried a search in the ysearch database with my five unusual markers and three of the most common for R1b (DYS19=14, DYS426=12, DYS388=12). There are 0 matches, among all haplogroups, other than myself.

I tried a search in the YHRD database, using 12 markers, including three of my unusual markers (DYS393=12, DYS385b=16, DYS437=16). There are 0 matches out of 48,122 haplotypes.
Logged
admixman
Member
**
Offline Offline

Posts: 8


« Reply #4 on: July 08, 2011, 02:53:17 AM »


Do you have a Ysearch ID?  I recommend creating one. You can do it from your FTDNA Y DNA Matches screen.

As RMS indicated, you could be classified as in the "11-13 Combo" group because of your unusual values at two fairly slow markers, DYS617=13 and DYS406s1>=11.
You apparently have a lot in common with "11-13" Group K, which is mostly the Kingston surnamed people.  Two people in 11-13 Group K tested positive for a new SNP named L513  so you might consider that test.  That would help determine your most recent branch on the Y DNA family tree. FTDNA calls it a haplotree.

Please join our project at http://www.familytreedna.com/public/R-L21-1113Combo/

Mikewww,
If you are the author of the L21+ spreadsheet, I say thank you and bravo. Excellent job.

I am entered in the ysearch database, and I have joined the the 11-13 Combo project.

I have ordered the L513 SNP from FTDNA.
Logged
OConnor
Old Hand
****
Offline Offline

Posts: 676


« Reply #5 on: July 08, 2011, 06:40:28 AM »

I am R-L21+...and...R-L159.2+

An L159.2+ common ancestor has been estimated to have lived perhaps 400AD.
I am 47/67 with Neal the L159.2 project administrator.

So I am not surprised that you think a gd of 20 (or so) is far reaching.

Logged

R1b1a2a1a1b4


R-DF13**(L21>DF13)
M42+, M45+, M526+, M74+, M89+, M9+, M94+, P108+, P128+, P131+, P132+, P133+, P134+, P135+, P136+, P138+, P139+, P14+, P140+, P141+, P143+, P145+, P146+, P148+, P149+, P151+, P157+, P158+, P159+, P160+, P161+, P163+, P166+, P187+, P207+, P224+, P226+, P228+, P229+, P230+, P231+, P232+, P233+, P234+, P235+, P236+, P237+, P238+, P239+, P242+, P243+, P244+, P245+, P280+, P281+, P282+, P283+, P284+, P285+, P286+, P294+, P295+, P297+, P305+, P310+, P311+, P312+, P316+, M173+, M269+, M343+, P312+, L21+, DF13+, M207+, P25+, L11+, L138+, L141+, L15+, L150+, L16+, L23+, L51+, L52+, M168+, M173+, M207+, M213+, M269+, M294+, M299+, M306+, M343+, P69+, P9.1+, P97+, PK1+, SRY10831.1+, L21+, L226-, M37-, M222-, L96-, L193-, L144-, P66-, SRY2627-, M222-, DF49-, L371-, DF41-, L513-, L555-, L1335-, L1406-, Z251-, L526-, L130-, L144-, L159.2-, L192.1-, L193-, L195-, L96-, DF21-, Z255-, DF23-, DF1-, Z253-, M37-, M65-, M73-, M18-, M126-, M153-, M160-, P66-

12 24 14 10 11 14 12 12 12 13 13 29 18


Mike Walsh
Guru
*****
Offline Offline

Posts: 2964


WWW
« Reply #6 on: July 08, 2011, 11:51:21 AM »

.... Alex Williamson calculates an age estimate for R-L21 of 3420±484 years, based on a mutation rate of 0.002 for all loci. ...

I don't advocate Alex W's method, I have no opinion one way or another.

However, I think we should keep our minds open that perhaps P312, L21, et al are younger than most will ever want to admit. Perhaps the tie to old Celtic language areas (and the known Celtic movements) is exactly what the geographic correlation appears to be.
Logged

R1b-L21>L513(DF1)>S6365>L705.2(&CTS11744,CTS6621)
rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #7 on: July 08, 2011, 09:31:11 PM »

I'm not surprised that 102 men have a distance of 20 or greater from the L21 modal. That is what I meant when I said that is not unusual. It isn't unusual. It isn't a majority or even a plurality, but 102 is a substantial number.

I'm sure that 4% at that range is about right for a haplogroup the age of R-L21.

But I don't think being one of those whose haplotype is greater than or equal to 20 from the modal is in itself anything all that significant. Last I checked, I was pretty far off the modal myself.

What is significant is belonging to an apparent clade or cluster, like the 11-13 combo group.
Logged

admixman
Member
**
Offline Offline

Posts: 8


« Reply #8 on: July 12, 2011, 07:36:27 PM »


But I don't think being one of those whose haplotype is greater than or equal to 20 from the modal is in itself anything all that significant. Last I checked, I was pretty far off the modal myself.

What is significant is belonging to an apparent clade or cluster, like the 11-13 combo group.


I follow what you mean by the value of a cluster versus a simple genetic distance.

What I am looking for in pursuing a “significantly off-modal” evaluation is a sense of how well a particular Y STR profile fits in with a group. There must be such an evaluation in some of the cluster allocations.

The number I am looking for should indicate how far away from the “center of mass” of a population an individual is. An outlier would be someone from an isolated population, an untested population, or possibly a population exposed to increased radiation. (Interestingly, my father worked on the Manhattan Project in the 1940’s, and my brother and I have a genetic distance of 1!)

I have come up with a number, which I call the weighted genetic distance (WGD). It can be used to determine a more meaningful value than the simple genetic distance, because it uses a specific mutation rate for each locus. The formula is:

WGD = ( ∑ {ABS(a_ref – a_ind) / mr_l })  * ( ∑ {mr_l} / n_l)

where:    
          a_ref = the allele value of the reference at a locus
          a_ind = the allele value of the individual at that locus
          mr_l = the mutation rate for that locus in mutations/generation
          n_l = number of loci
          { } = array values

In Excel this translates to:

{=SUM(ABS($A$3:$BO$3-A4:BO4)/$A$2:$BO$2)*(SUM($A$2:$BO$2)/67)}

where:   
          the 67 loci names are in row 1, columns A to BO
          the 67 mutation rates are in row 2, columns A to BO
          the 67 alleles of the reference population are in row 3, columns A to BO
          the 67 alleles of the individual are in row 4, columns A to BO

I am in the process of building a spreadsheet with all of the R-L21 project data to see how usable this WGD is.

I would appreciate comments on this. Am I reinventing the wheel? Have others already developed this approach?
Logged
Mike Walsh
Guru
*****
Offline Offline

Posts: 2964


WWW
« Reply #9 on: July 12, 2011, 10:39:30 PM »

...
What I am looking for in pursuing a “significantly off-modal” evaluation is a sense of how well a particular Y STR profile fits in with a group. There must be such an evaluation in some of the cluster allocations.

The number I am looking for should indicate how far away from the “center of mass” of a population an individual is. An outlier would be someone from an isolated population, an untested population, or possibly a population exposed to increased radiation. (Interestingly, my father worked on the Manhattan Project in the 1940’s, and my brother and I have a genetic distance of 1!)
....
I have tried this and have what I call "normalized" Genetic Distance and "normalized" variance based on STR mutation rates.

I've asked several scientists who are involved in genetic genealogy. Everytime I ask I'm put back in my place in relation to the fact this is insiignificant over a number of markers. I don't know, but the right place to ask your question is on Rootsweb where you'll have Harvard, Stanford, MIT and National Science Board guys (and maybe even a Russin biochemist) answer.

http://archiver.rootsweb.ancestry.com/th/index/GENEALOGY-DNA
Logged

R1b-L21>L513(DF1)>S6365>L705.2(&CTS11744,CTS6621)
admixman
Member
**
Offline Offline

Posts: 8


« Reply #10 on: July 13, 2011, 01:48:18 AM »


I have tried this and have what I call "normalized" Genetic Distance and "normalized" variance based on STR mutation rates.

I've asked several scientists who are involved in genetic genealogy. Everytime I ask I'm put back in my place in relation to the fact this is insiignificant over a number of markers. I don't know, but the right place to ask your question is on Rootsweb where you'll have Harvard, Stanford, MIT and National Science Board guys (and maybe even a Russin biochemist) answer.

http://archiver.rootsweb.ancestry.com/th/index/GENEALOGY-DNA

I like it here for now.

I am just beginning to explore the usefulness of a weighted genetic distance (or normalized genetic distance) with my recently built R-L21 spreadsheet.

Already I have found that it is perhaps more useful to use one person as the reference (rather than the L21 modal) and see how all the other individuals compare to that one person.

The WGD may also be useful for finding significant anomalies or errors in the data.

For instance, there are only three cases in the L21 project where the WGD exceeds 100, but one really stands out. Mr. Whalen (f55192, descendant of John Phalen) has a modest GD of 19 off the modal, but a whopping 409 for the WGD. Why? Because he has DYS472=9 (instead of 8), DYS578=10 (instead of 9), and DYS413a=16 (7 steps off the modal of 23). There seems to be no other case of DYS472=9 in anyone in any R haplogroup. This is truly an exceptionally rare haplotype.

However, upon further examination, we see that there is a Mr. Whelan (f83115, descendant of Hugh Whalin), spelled with an exchange of vowels, who has the same origin (Ireland, Co. Laois). There is only a GD of 4 between them. They clearly should be related, especially since he shares the rare DYS578=10 and the rare DYS413a=16. However he carries the nearly universal DYS472=8.

This case points to a testing error, a data entry error, or an anomaly that is not consistent with the mutation rates.
Logged
Pages: [1] Go Up Print 
« previous next »
Jump to:  


SEO light theme by © Mustang forums. Powered by SMF 1.1.13 | SMF © 2006-2011, Simple Machines LLC

Page created in 0.112 seconds with 19 queries.