PDA

View Full Version : What thresholds can be used to confirm more distant ancestry



psampson
08-08-2017, 09:28 PM
This section on the ISOGG wiki page on inheritence by descent caught my eye and I'm sure it's a quote with which the forum is familiar
"for the most part IBD blocks longer than 4 cM come from 500 to 1,500 years ago, and blocks longer than 10 cM are within the last 500 years"

I would like to know what to look for when dealing with segment blocks, what do IBD and IBC blocks look like and can phasing a parental kit confirm a genetic link.

The family group I've been researching dates to around 1300 in Suffolk. In the 1400's two branches were quite literally in the next field in North Norfolk to my direct paternal line, and the branch from the 1600's appears in a Gedmatch members gedcom file. With my kit, my sisters kit and my Dad's phased kit I dropped the thresholds to 3cM and 250 SNP's to maximise results and hopefully illustrate my question about segment blocks and came up with the following results.

This Gedmatch member has four kits (I've posted the first two), presumably from close family members and all three of my kits match all four of his kits. Are they IBD or IBC?

My kit
Chr Start Location End Location Centimorgans (cM) SNPs
1 2,270,521 3,302,774 3.4 267
1 17,748,976 18,640,364 3.3 377
4 8,172,699 9,905,396 3.4 351
4 144,715,927 148,600,255 3.2 564
5 174,236,006 175,869,122 3.7 446
5 175,908,000 178,050,226 3.6 492
11 125,947,716 128,021,449 3.7 548
14 88,774,473 89,712,654 3.0 255
15 24,463,770 25,392,155 3.3 251
16 11,779,901 13,225,980 3.9 485
19 54,426,874 55,797,960 3.6 316
20 58,224,844 58,998,640 4.1 254
22 33,818,632 35,125,530 3.0 354

Sister's kit
Chr Start Location End Location Centimorgans (cM) SNPs
1 17,762,277 18,640,378 3.2 392
2 29,582,835 31,421,433 3.3 635
2 193,473,278 199,077,747 3.5 803
5 8,537,044 10,634,830 5.4 696
18 5,976,416 6,858,925 3.9 280
18 59,031,542 60,691,045 3.1 478
22 20,861,074 22,540,725 4.6 283
22 23,933,018 24,936,256 3.2 292
22 33,818,632 35,125,530 3.0 360

Dad's phased kit
Chr Start Location End Location Centimorgans (cM) SNPs
4 8,172,699 9,905,396 3.4 310
11 125,947,716 128,021,449 3.7 512




On this kit, the SNP count is slightly different
My kit
Chr Start Location End Location Centimorgans (cM) SNPs
1 2,270,521 3,302,774 3.4 262
1 17,748,976 18,640,364 3.3 376
4 8,172,699 9,905,396 3.4 348
4 144,715,927 148,600,255 3.2 562
5 174,236,006 175,869,122 3.7 445
5 175,908,000 178,050,226 3.6 490
11 125,947,716 128,021,449 3.7 542
14 88,774,473 89,712,654 3.0 254
15 24,463,770 25,392,155 3.3 251
16 11,779,901 13,225,980 3.9 484
19 54,426,874 55,797,960 3.6 314
20 58,224,844 58,998,640 4.1 253
22 33,818,632 35,125,530 3.0 353


Sister's kit
Chr Start Location End Location Centimorgans (cM) SNPs
1 17,762,277 18,640,378 3.2 396
2 29,582,835 31,421,433 3.3 645
2 193,473,278 199,077,747 3.5 817
5 8,537,044 10,634,830 5.4 702
18 59,031,542 60,691,045 3.1 488
22 23,933,018 24,936,256 3.2 306

Dad's phased kit
Chr Start Location End Location Centimorgans (cM) SNPs
4 8,172,699 9,905,396 3.4 307
11 125,947,716 128,021,449 3.7 506


There are 15 kits with which I, my sister & Dad share matching segments, two of them share 8.8cM & 10.1cM with me & Dad (8.7cM & 9cM)

Saetro
08-10-2017, 08:11 AM
Presumably the page you looked at was https://isogg.org/wiki/Identical_by_descent
From what you say above, you appear to be making thoughtful, considered statements.
Other people may need to be much more careful, and the first few of my comments below are aimed at them.

There are at least a couple of problems with small segments, such as -
1) Was this segment handed down from an ancestor, or does it consist of bits that have just come together by chance and look like a single segment?
The smaller the segment, the more likely this is.
2) If small segments come from a long time ago, does your proven pedigree go back far enough?
There are two statistics often quoted: as a long standing European, you will be descended from everyone who was in Europe before 1350; and about 1/3 of English can be shown to be descended from Edward III.
So, having the paperwork is quite another thing.

Having said that, I have found a couple of matches at 11C to 13C, one in a line that goes back to Henry I with some assurance, so can be taken further.
But that is only two of a couple of thousand of ancestors at that distance (circa 1500), and the number rises exponentially, so becomes much bigger quite soon.
And the matching segments were >8cM so had a better chance of being real.

Ultimately a segment is IBD if it really does link to a person.
That is going to be decided not just on segment length, but together with other evidence usually from documentary sources.

There is lots about this topic from people with a heap of experience.
Two say that small segments can be useful, IFF (if and only if) you have other suitable evidence.
Jim Bartlett's blog segmentology.org talks about this here and there.
Dave Hamm also has a blog, but most of his comments (often with Jim's comments) on this topic are here and there at http://archiver.rootsweb.ancestry.com/th/read/GENEALOGY-DNA/