RootsWeb.com Mailing Lists
Total: 8/8
    1. [AUTOSOMAL-DNA] Question about a match
    2. Christina Hunt
    3. My husband's brother recently got a decent match (13.3 cm) on Chromosome 1. It was of interest because many of her ancestors and my husband's Irish ancestor are from the same small town. My husband and his cousin didn't show up at all. Just wondering if my thinking is right. On Chromosome 1 there are two segments that start or end with the same number with the brothers ie 102,289,362. Is this because the chromosome has split up further? Does the fact that I am only finding one match meet the 7 cm threshold that the common ancestor is farther back than the 5+ generations suggested. I have researched this person's tree as far back as 5 generations with nothing other than location seeming to match. Brother one 1 3,645,052 4,379,687 2.2 233 1 102,289,362 105,725,458 2.8 541 << 1 160,318,984 162,630,894 3.8 611 1 242,594,868 244,106,113 3.0 275 << 1 245,630,414 246,429,058 2.2 254 Brother two 1 3,645,052 4,379,687 2.2 230 1 102,289,362 105,720,186 2.8 531 << 1 160,420,036 162,722,216 3.7 599 1 167,532,944 168,569,112 2.0 288 1 216,240,301 230,862,034 13.3 3,472 1 242,599,000 244,106,113 3.0 266 << 1 245,630,414 246,429,058 2.2 253 For Brother two Largest segment = 13.3 cM Total Half-Match segments (HIR) = 144.4 cM (4.027 Pct) 52 shared segments found for this comparison. 613979 SNPs used for this comparison. 55.774 Pct SNPs are full identical Thanks, Chris

    01/20/2019 11:36:41
    1. [AUTOSOMAL-DNA] Re: Question about a match
    2. Tim Janzen
    3. Dear Chris, There is a high probability that many or all of the HIRs under 4 cMs are false matches. I suggest that you ignore those unless you have proof that they are IBD. Just because 2 siblings share an HIR with a match that starts and/or ends on the same position doesn't necessarily indicate that the HIR in question is IBD. I have definitely run into situations in my family where two crossovers have occurred in close proximity to each other on the same chromosome which has resulted in a gap between 2 IBD segments that were passed down from the same ancestor. However, the IBD segments were over 7 cMs in those cases. Sincerely, Tim Janzen -----Original Message----- From: Christina Hunt [mailto:[email protected]] Sent: Sunday, January 20, 2019 10:37 AM To: [email protected] Subject: [AUTOSOMAL-DNA] Question about a match My husband's brother recently got a decent match (13.3 cm) on Chromosome 1. It was of interest because many of her ancestors and my husband's Irish ancestor are from the same small town. My husband and his cousin didn't show up at all. Just wondering if my thinking is right. On Chromosome 1 there are two segments that start or end with the same number with the brothers ie 102,289,362. Is this because the chromosome has split up further? Does the fact that I am only finding one match meet the 7 cm threshold that the common ancestor is farther back than the 5+ generations suggested. Thanks, Chris

    01/20/2019 02:55:13
    1. [AUTOSOMAL-DNA] Re: Question about a match
    2. Christina Finn Hunt
    3. Thanks Tim. On Sun, 20 Jan 2019, 4:55 PM Tim Janzen <[email protected] wrote: > Dear Chris, > There is a high probability that many or all of the HIRs under 4 cMs are > false matches. I suggest that you ignore those unless you have proof that > they are IBD. Just because 2 siblings share an HIR with a match that > starts > and/or ends on the same position doesn't necessarily indicate that the HIR > in question is IBD. I have definitely run into situations in my family > where two crossovers have occurred in close proximity to each other on the > same chromosome which has resulted in a gap between 2 IBD segments that > were > passed down from the same ancestor. However, the IBD segments were over 7 > cMs in those cases. > Sincerely, > Tim Janzen > > -----Original Message----- > From: Christina Hunt [mailto:[email protected]] > Sent: Sunday, January 20, 2019 10:37 AM > To: [email protected] > Subject: [AUTOSOMAL-DNA] Question about a match > > My husband's brother recently got a decent match (13.3 cm) on Chromosome > 1. It was of interest because many of her ancestors and my husband's > Irish ancestor are from the same small town. > > My husband and his cousin didn't show up at all. > Just wondering if my thinking is right. On Chromosome 1 there are two > segments that start or end with the same number with the brothers ie > 102,289,362. > > Is this because the chromosome has split up further? Does the fact that > I am only finding one match meet the 7 cm threshold that the common > ancestor is farther back than the 5+ generations suggested. > > Thanks, > Chris > > _______________________________________________ > > ______________________________ > For answers to Frequently Asked Questions about mailing lists, please see: > http://dgmweb.net/MailingListFAQs.html > > _______________________________________________ > Email preferences: http://bit.ly/rootswebpref > Unsubscribe > https://lists.rootsweb.com/postorius/lists/[email protected] > Privacy Statement: https://ancstry.me/2JWBOdY Terms and Conditions: > https://ancstry.me/2HDBym9 > Rootsweb Blog: http://rootsweb.blog > RootsWeb is funded and supported by Ancestry.com and our loyal RootsWeb > community >

    01/20/2019 04:08:55
    1. [AUTOSOMAL-DNA] Re: Question about a match
    2. McDonald, J Douglas
    3. A different question. What is the probability that a 4 or 5 or ... 10 CM HIR segment is really IBD, from a matching person that has a purported (on paper) matching pair of ancestors n generations back, and m other surname (not exact person) matches also n generations back. One of course has to specify the commonness of the surname list. This is a non-triangulated segment. This sort of statistics make my head spin. Our University has a Statistical Help Service for things "sort of" like this. One talks to a grad student flunkie first and if they can't help, one gets to ask a real professor. I've tried questions "like" this and in fact this exact one. They give up ... I think they can only answer exact canned problems! But there has to be an Answer! Oh yes ... one has to multiply in one minus the probability that the paper match is a bogus Ancestry.com wrong paper trail ... but this is just a general static multiplier. An answer to this would be a miracle. Doug McDonald ________________________________ From: Christina Finn Hunt <[email protected]> Sent: Sunday, January 20, 2019 5:08 PM To: [email protected] Subject: [AUTOSOMAL-DNA] Re: Question about a match Thanks Tim. On Sun, 20 Jan 2019, 4:55 PM Tim Janzen <[email protected] wrote: > Dear Chris, > There is a high probability that many or all of the HIRs under 4 cMs are > false matches. I suggest that you ignore those unless you have proof that > they are IBD. Just because 2 siblings share an HIR with a match that > starts > and/or ends on the same position doesn't necessarily indicate that the HIR > in question is IBD. I have definitely run into situations in my family > where two crossovers have occurred in close proximity to each other on the > same chromosome which has resulted in a gap between 2 IBD segments that > were > passed down from the same ancestor. However, the IBD segments were over 7 > cMs in those cases. > Sincerely, > Tim Janzen > > -----Original Message----- > From: Christina Hunt [mailto:[email protected]] > Sent: Sunday, January 20, 2019 10:37 AM > To: [email protected] > Subject: [AUTOSOMAL-DNA] Question about a match > > My husband's brother recently got a decent match (13.3 cm) on Chromosome > 1. It was of interest because many of her ancestors and my husband's > Irish ancestor are from the same small town. > > My husband and his cousin didn't show up at all. > Just wondering if my thinking is right. On Chromosome 1 there are two > segments that start or end with the same number with the brothers ie > 102,289,362. > > Is this because the chromosome has split up further? Does the fact that > I am only finding one match meet the 7 cm threshold that the common > ancestor is farther back than the 5+ generations suggested. > > Thanks, > Chris > > _______________________________________________ > > ______________________________ > For answers to Frequently Asked Questions about mailing lists, please see: > http://dgmweb.net/MailingListFAQs.html > > _______________________________________________ > Email preferences: http://bit.ly/rootswebpref > Unsubscribe > https://lists.rootsweb.com/postorius/lists/[email protected] > Privacy Statement: https://ancstry.me/2JWBOdY Terms and Conditions: > https://ancstry.me/2HDBym9 > Rootsweb Blog: http://rootsweb.blog > RootsWeb is funded and supported by Ancestry.com and our loyal RootsWeb > community > _______________________________________________ ______________________________ For answers to Frequently Asked Questions about mailing lists, please see: http://dgmweb.net/MailingListFAQs.html _______________________________________________ Email preferences: http://bit.ly/rootswebpref Unsubscribe https://lists.rootsweb.com/postorius/lists/[email protected] Privacy Statement: https://ancstry.me/2JWBOdY Terms and Conditions: https://ancstry.me/2HDBym9 Rootsweb Blog: http://rootsweb.blog RootsWeb is funded and supported by Ancestry.com and our loyal RootsWeb community

    01/20/2019 05:47:26
    1. [AUTOSOMAL-DNA] Re: Question about a match
    2. Tim Janzen
    3. Dear Doug, This really is best broken up into two questions: 1. What is the probability that an HIR of n length in cMs and containing m SNPs is IBD? This has been fairly well addressed in my opinion by analysis that John Walden, others, and I have done. See https://isogg.org/wiki/Identical_by_descent for a summary of what John and I have done. The bottom line is that the 50-50 split between IBD and false matches is about 7 cMs. I continue to believe that no one has ever demonstrated that an HIR over 15 cMs containing at least 2000 SNPs was a false match. The number of SNPs is an important variable in the entire issue of whether an HIR is IBD or is a false match. HIRs that contain less than 500 steps are frequently false matches. HIRs that contain between 500 and 1000 SNPs are sometimes false matches. HIRs that contain between 1000 and 1500 SNPs are sometimes false matches not all that frequently. 2. What is the probability than an IBD segment truly came from a shared ancestor that you have identified n number of generations back in time? The answer to this question depends on a number of significant variables including the following: 1. The depth in terms of the number generations back in time the pedigree charts have been accurately established for the 2 people sharing the HIR. 2. The relative degree of endogamy in the ancestries of the 2 people sharing the HIR. In my experience, relatively few people have their pedigree charts accurately trace back more than about 8 generations on all lines. This complicates things substantially. We also know from the simulations that Speed and Balding did (see https://isogg.org/wiki/Identical_by_descent) that many IBD segments over 10 cMs involve a shared ancestor 20 generations or more back in time. You therefore have to be fairly careful when you are attributing DNA segments to specific ancestors. Ideally, you do extensive chromosome mapping using data from 1st to 10th cousins to "walk an IBD segment back up your family three". Triangulation where you have 3 or more people who share the same ancestor and the same segment n number of generations back in time helps provide evidence, but even in such situations I suggest caution if one or more of the matches doesn't have the segment accurately mapped back to an ancestor relatively closely related to the shared ancestor n number of generations back. Sincerely, Tim -----Original Message----- From: McDonald, J Douglas [mailto:[email protected]] Sent: Sunday, January 20, 2019 4:47 PM To: [email protected] Subject: [AUTOSOMAL-DNA] Re: Question about a match A different question. What is the probability that a 4 or 5 or ... 10 CM HIR segment is really IBD, from a matching person that has a purported (on paper) matching pair of ancestors n generations back, and m other surname (not exact person) matches also n generations back. One of course has to specify the commonness of the surname list. This is a non-triangulated segment. This sort of statistics make my head spin. Our University has a Statistical Help Service for things "sort of" like this. One talks to a grad student flunkie first and if they can't help, one gets to ask a real professor. I've tried questions "like" this and in fact this exact one. They give up ... I think they can only answer exact canned problems! But there has to be an Answer! Oh yes ... one has to multiply in one minus the probability that the paper match is a bogus Ancestry.com wrong paper trail ... but this is just a general static multiplier. An answer to this would be a miracle. Doug McDonald

    01/20/2019 07:01:03
    1. [AUTOSOMAL-DNA] Re: Question about a match
    2. Eric S Johnson
    3. > What is the probability that a 4 or 5 or ... 10 CM HIR segment is really IBD, > from a matching person that has a purported (on paper) matching pair of > ancestors > n generations back, and m other surname (not exact person) matches also n > generations back. One of course has to specify the commonness of the > surname list. > This is a non-triangulated segment. Seems to me the "that has a purported matching pair of ..." qualification is irrelevant; whether there's a known paper-trail connection doesn't affect the probability. But, in the end, all that's relevant is whether it triangulates. Don't we find that *all* chromosomal addresses eventually have two TGs at them, and Doug's HIR thereupon either does, or does not, fall into one of those TGs? In which case the question of a small HIR being IBD becomes irrelevant ... or, at any rate, academic ... no?

    01/20/2019 09:04:47
    1. [AUTOSOMAL-DNA] Re: Question about a match
    2. Tim Janzen
    3. Since we have two of each autosomal chromosome then we have two triangulated groups for any region of a specific chromosome. If an HIR represents an IBD segment then it will be in a triangulated group. However, the smaller the IBD segment is, the lower the probability that we will be able to determine the shared ancestor. Tim -----Original Message----- From: Eric S Johnson [mailto:[email protected]] Sent: Sunday, January 20, 2019 8:05 PM To: [email protected] Subject: [AUTOSOMAL-DNA] Re: Question about a match Seems to me the "that has a purported matching pair of ..." qualification is irrelevant; whether there's a known paper-trail connection doesn't affect the probability. But, in the end, all that's relevant is whether it triangulates. Don't we find that *all* chromosomal addresses eventually have two TGs at them, and Doug's HIR thereupon either does, or does not, fall into one of those TGs? In which case the question of a small HIR being IBD becomes irrelevant ... or, at any rate, academic ... no?

    01/20/2019 09:44:48
    1. [AUTOSOMAL-DNA] Re: Question about a match
    2. McDonald, J Douglas
    3. Most matches cannot triangulate. A match cannot triangulate without chromosome level detail. Anceastry.com, where most matches with paper trails are, does not provide chromosome level detail, and requests to the match for the raw data almost always fail. I've said before, I have a 33% success rate for very close cousins, and zero outside that range. That's why I asked my question. Whether there is a paper trail most certainly DOES affect the probability of IBD. The Rev. Bayes long ago explained that. I can do extremely rough guesses using his method in my head, getting numbers combined right needs expertise. Doug -----Original Message----- From: Tim Janzen <[email protected]> Sent: Sunday, January 20, 2019 10:45 PM To: [email protected] Subject: [AUTOSOMAL-DNA] Re: Question about a match Since we have two of each autosomal chromosome then we have two triangulated groups for any region of a specific chromosome. If an HIR represents an IBD segment then it will be in a triangulated group. However, the smaller the IBD segment is, the lower the probability that we will be able to determine the shared ancestor. Tim

    01/21/2019 08:01:55