Need explanation, PLZ

[Ginny-Kay]

There are 3 groups (A,B,C) who have tested 37 markers.
For DYS464b, Groups A and B have 17 and Group C has 16.
For DYS456, Groups A and C have 16 and Group B has 17.
Would someone please translate this into probabilities of relationship? eg grp A and B were probably brothers and C a cousin, or whatever!!
Thank you for help with this. Trying to break down some long-standing brick walls!!
Ginny-Kay

[nathanm]

There are 3 groups (A,B,C) who have tested 37 markers.
For DYS464b, Groups A and B have 17 and Group C has 16.
For DYS456, Groups A and C have 16 and Group B has 17.
Would someone please translate this into probabilities of relationship? eg grp A and B were probably brothers and C a cousin, or whatever!!
Thank you for help with this. Trying to break down some long-standing brick walls!!
Ginny-Kay

Are those the only differences in 37 markers? If so, A and B have a GD of 1, A and C have a GD of 1, and B and C have a GD of 2. The easiest explanation is that they all share a common ancestor, who had the same haplotype as A. Then two separate descendants had different mutations, one on DYS456, whose descendants are in group B, and the other on DYS464b, whose descendants are in group C.

[Ginny-Kay]

And...

Oh, dear, wrote out more questions and lost them. Thank you for your reply.

Found a single person (group D) also GD1 from A and GD2 from B & C. Are we talking descendents from 4 brothers, one line which didn't mutate? Or, are we talking four lines from one father (only one line not mutating?), or what?

I have a paper trail for 3 of the 4 descendents in Group A which equals 7 generations without a common ancestor (all 37 markers the same, one different surname). So how far back are we looking?

Also, if the other 3 groups have a genetic distance of 2 from each other, how far back do those go to a common ancestor?

Sorry for the questions and appreciative of the help. I'm too old to have had any genetics in school except red, white and pink petunias.

[nathanm]

Oh, dear, wrote out more questions and lost them. Thank you for your reply.

Found a single person (group D) also GD1 from A and GD2 from B & C. Are we talking descendents from 4 brothers, one line which didn't mutate? Or, are we talking four lines from one father (only one line not mutating?), or what?

You do realize that's two ways of saying the exact same thing, right? It's definitely a possibility. Are they all in the same haplogroup? If not, we can rule out a real match. Are these groups part of a surname DNA project? If so, the project administrator may have more information, perhaps about matches who've confirmed how they're related.

I have a paper trail for 3 of the 4 descendents in Group A which equals 7 generations without a common ancestor (all 37 markers the same, one different surname). So how far back are we looking?

Also, if the other 3 groups have a genetic distance of 2 from each other, how far back do those go to a common ancestor?

Sorry for the questions and appreciative of the help. I'm too old to have had any genetics in school except red, white and pink petunias.

According to Dr McDonald's TMRCA Calculator, here are the probabilities for the matches, using the default overall average mutation rate for 37 markers (0.0033):

• 37/37 = 50% chance of a common ancestor within 3 generations, 90% within 10
• 36/37 = 50% within 7 generations, 90% within 16
• 35/37 = 50% within 11 generations, 90% within 22

However, for the groups with a GD of 1, it's easy to substitute the actual observed mutation rate for the marker that differs, which should make the estimates more accurate. The probabilities may be significantly better, since the markers they differ on mutate faster than the overall average: DYS464b at 0.00566/generation and DYS456 at 0.00735 (more than twice as fast).

• Groups A and C (DYS464b): 36/37 = 50% within 4 generations, 90% within 9
• Groups A and B (DYS456): 36/37 = 50% within 3 generations, 90% within 7

So the odds these groups are related are better than the ones listed above. Which marker does D differ on?

Using a 30 year average generation length, a perfect match has even odds of a common ancestor within about 90 years, and pretty good odds within 300. The others would be similar. Did the ancestors live in the same areas? That would be a good indicator it's plausible. Regarding the different surname, is there any ambiguity in the paper trail that might suggest an NPE? Have any tested beyond 37 markers? The more they test and match on, the better you can refine the estimate to a common ancestor.

[M.O'Connor]

Ginny Kay

Do you have y-search profiles for each of these sequences? It might be easier to actually compare the sequences that way.

[Ginny-Kay]

Reply

Again, I wrote out quite a lengthy reply and then was not allowed to post as it said I wasn't logged in. Perhaps, there is not quite enough time allowed before we are dumped?

Group D has a 12 on marker 385a and ABC have 11.

ABC are on a surname project whose administrator is new. I think one of the group A members has asked, but did not get answers. I have been in contact with all ABC members and trying with D (thru YSearch). All appear to have the same Haplotype (at least as far as it has been tested). Most are inferred:
1b1a2a and the tested ones are 1b1a2a1a1b5a.

Paper trail for Group A goes back to MD/VA? 1730s. Definitely prove them in Granville Co (now Caswell), Anson Co, NC around 1750. They seem to have traveled south to SC, GA, ALA, Miss, LA/ARK, TX.

Paper trail for Group B goes back to GA around 1800. Families they married into go back to Va/NC. They may have come south down the coast from Tidewater VA to Chowan Co, NC.

Paper trail for Groups C&D seem to go back to a common ancestor in Henrico Co, VA 1680-1720. They also moved so to Granville Co, NC, but appear to have then gone to TN (at least some of them) before moving to TX.

Thank you for your help.

[JohnLloydScharf]

Your best way to understand this is to contact the project administrator to see which group you best fit into. I would base groupings on known ancestors and slowest mutation markers, but that would be a judgement call.

There are 3 groups (A,B,C) who have tested 37 markers.
For DYS464b, Groups A and B have 17 and Group C has 16.
For DYS456, Groups A and C have 16 and Group B has 17.
Would someone please translate this into probabilities of relationship? eg grp A and B were probably brothers and C a cousin, or whatever!!
Thank you for help with this. Trying to break down some long-standing brick walls!!
Ginny-Kay

[k.o.gran]

Hi Ginny-Kay.

As M. O'Connor said, it will probably be easier to get help if you import the results into ysearch.org and post the IDs.

[nathanm]

Again, I wrote out quite a lengthy reply and then was not allowed to post as it said I wasn't logged in. Perhaps, there is not quite enough time allowed before we are dumped?

Group D has a 12 on marker 385a and ABC have 11.

ABC are on a surname project whose administrator is new. I think one of the group A members has asked, but did not get answers. I have been in contact with all ABC members and trying with D (thru YSearch). All appear to have the same Haplotype (at least as far as it has been tested). Most are inferred:
1b1a2a and the tested ones are 1b1a2a1a1b5a.

Paper trail for Group A goes back to MD/VA? 1730s. Definitely prove them in Granville Co (now Caswell), Anson Co, NC around 1750. They seem to have traveled south to SC, GA, ALA, Miss, LA/ARK, TX.

Paper trail for Group B goes back to GA around 1800. Families they married into go back to Va/NC. They may have come south down the coast from Tidewater VA to Chowan Co, NC.

Paper trail for Groups C&D seem to go back to a common ancestor in Henrico Co, VA 1680-1720. They also moved so to Granville Co, NC, but appear to have then gone to TN (at least some of them) before moving to TX.

Thank you for your help.

DYS385a is a slower mutating marker, so the MRCA between D and A comes out to 50% within 10 generations, and 90% within 23. However, it's still possible they share a common ancestor; those are just probabilities. If the various paper trails are solid, I'd try to determine who was in the right time and place to figure out how they might fit into a hypothetical family tree. It sounds like they all have North Carolina in common.

[JohnLloydScharf]

By no means is 385a or 385b a "slower mutating marker." It's rate is 0.00226. That is slower than the typical average used by software and for MRCA. These are documented slower moving markers:
DYS426 GTT 10-12 0.00009
DYS454 AAAT 10-12 0.00016
DYS455 AAAT 8-12 0.00016
DYS388 ATT 10-16 0.00022
DYS392 TAT 6-17 0.00052

DYS438 TTTTC 6-14 0.00055
DYS393 AGAT 9-17 0.00076
DYS393 is AKA DYS395
DYS437 TCTA 13-17 0.00099
YCAIIa 0.00123
YCAIIb 0.00123

DYS459 0.00132
DYS448AGAGAT17-24 0.00135
DYS394 TAGA 10-19 0.00151
DYS394 is AKA DYS19.
DYS389i 0.00186

If you only had 385a as a missed marker it would depend on how many markers were tested, regarding the MRCA probability. If it is the only one in a test of 111 markers, you are almost twins. If it is the only one in a test of 12 markers, you are excluded as likely for 100 generations.

DYS385a is a slower mutating marker, so the MRCA between D and A comes out to 50% within 10 generations, and 90% within 23. However, it's still possible they share a common ancestor; those are just probabilities. If the various paper trails are solid, I'd try to determine who was in the right time and place to figure out how they might fit into a hypothetical family tree. It sounds like they all have North Carolina in common.

[Ginny-Kay]

Y-search

Thank all of you for your replies. All these groups are on Y-search, and all but Group D are on a Surname Project which lumps them together as Subgroup 5, so that's not very helpful. I have scoured the records for nearly 45 years and I still can not tie the three members of Group A with which I am most familiar together. Nor can I fit A with B,C,D from Virginia to present. There are a few possible places left where a common ancestor might be found for Group A, but it would have to be within GroupC,D's family, and, if I understand earlier replies, that is not likely. So, my guess is I will have to go back across the Atlantic to find the common ancestor.
Thank all of you again for your help. Guess I will chew on this for awhile.
Ginny-Kay

[JohnLloydScharf]

Ginny-Kay, let's start from the beginning. We need a surname.Your ABCs are confusing.

You have talked about migrations that match most of the population of the US since the end of the 17th Century in the US. Henrico, County seems to have been a jumping off place for immigrants who rode the "William and Mary" out from London who came from many parts of Western Europe. Parts of Henrico became Goochland by when Sir Gooch was the only one of any real nobility came to oversea Virginia.

Many immigrants' families followed the same routes from Maryland and Virginia to North Carolina, South Carolina, Georgia, Arkansas, and Texas, if they did not go to Kentucky or Oklahoma because of mixing with Native Americans. Your surname could be Childress from just THAT description.

[nathanm]

Sometimes, I don't know why I even bother...

By no means is 385a or 385b a "slower mutating marker." It's rate is 0.00226. That is slower than the typical average used by software and for MRCA.

By no means? You contradict yourself two sentences later. It's slower than the overall average mutation rate for 37 markers (0.0033), by almost a third. I never said it was the slowest marker, just slower. And that made a significant difference in the MRCA calculation.

If you only had 385a as a missed marker it would depend on how many markers were tested, regarding the MRCA probability. If it is the only one in a test of 111 markers, you are almost twins. If it is the only one in a test of 12 markers, you are excluded as likely for 100 generations.

We do know how many markers were tested and matched, 36 out of 37.

[Ginny-Kay]

Continuing saga

All the people involved are on Ysearch and all but one are on a FT Surname project. I have been working this surname for 44 years. I do not understand the suggestion to post the IDs of the people on Ysearch.

I believe Nathan M correctly interrpreted the possibilities of the relationship between the man in Group D and the Group A people at 37 markers. At least from how I understand what JLSharf said, the math seems to be close (I'm not a roaring star in math). In addition, I understand that any of these mutations could happen at any time, so the guesses are just that, educated guesses, probablilites, possibilities and odds.

[Ginny-Kay]

And...

Yes I understand about Henrico, and jumping off places. THe surname is Robinson/Robertson and the other 41 variations of same. I can see the movement patterns, but how much is because these 3-4 groups of people knew they were kin and thus traveled together, and how much is because everyone was receiving land grants in the same areas at the same times is more than I can presently guess.

I am simply trying to get some hints from the DNA as to what directions to search next and what I can expect to find.