PDA

View Full Version : IBD Or IBS?

J Man
11-28-2014, 11:01 PM
How large does a chromosme segment match have to be to be considered a true IBD segment and not an IBS segment generally? Is it around 7.5cM?

ArmandoR1b
11-28-2014, 11:24 PM
IBS can be as large as 12.9 cM See http://www.jogg.info/72/files/Turner.htm

AJL
11-29-2014, 12:13 AM
As Armando alludes to, it's really a question of probabilities. From comparing multiple kits of my relatives I can see there are some surprisingly short IBD segments (around 3 cM) but the odds of that in two random people must be rather low. On the high end, meanwhile, some segments above 10 cM will be IBS, but not that many.

It's a tradeoff between a wild goose chase, and not enough matches. I've found that sticking to segments above, say, 9-10 and which only appear in two related people works for me.

J Man
11-29-2014, 03:05 AM
As Armando alludes to, it's really a question of probabilities. From comparing multiple kits of my relatives I can see there are some surprisingly short IBD segments (around 3 cM) but the odds of that in two random people must be rather low. On the high end, meanwhile, some segments above 10 cM will be IBS, but not that many.

It's a tradeoff between a wild goose chase, and not enough matches. I've found that sticking to segments above, say, 9-10 and which only appear in two related people works for me.

Okay thank you both for the information. What exactly do you mean by this? ''I've found that sticking to segments above, say, 9-10 and which only appear in two related people works for me.'' Do you mean that the same segment will appear in say you and your parent with the person you match? That is your parent will share the same segment with that match as you do?

AJL
11-29-2014, 04:23 PM
Okay thank you both for the information. What exactly do you mean by this? ''I've found that sticking to segments above, say, 9-10 and which only appear in two related people works for me.'' Do you mean that the same segment will appear in say you and your parent with the person you match? That is your parent will share the same segment with that match as you do?

I have some segments around 9-10 cM that appear in me but in neither parent -- I generally disregard those. If a segment appears in either parent at 9-10 cM, or in me plus a parent, I will likely pursue it.

J Man
11-29-2014, 04:37 PM
I have some segments around 9-10 cM that appear in me but in neither parent -- I generally disregard those. If a segment appears in either parent at 9-10 cM, or in me plus a parent, I will likely pursue it.

Okay so say I have a 15 cM segment match with a person and my father also shares that same segment match with that person at the same location on a chromosome then that segment match is likely truly IBD correct?

ArmandoR1b
11-29-2014, 05:11 PM
Using parents, as AJL is doing, is really how we are supposed to go about determining IBS vs IBD but not everyone has the luxury of being able to test their parents so the next thing to do is to see if your match is able to have their parents tested and see if the match still exists. If not it is IBS. If it does you still have to try triangulation with other people. Gedmatch has Matching Segment Search and Triangulation in their Tier 1 Utilities which you can get with a one month non-recurring payment of \$10. Genome Mate can also help but it doesn't tell you if your matches match each other. You can do that with Gedmatch for the matches that have uploaded to Gedmatch.

jeanL
12-05-2014, 01:38 PM
Check this out:

Today somebody added me, with whom I share 31 Mb, yet when I checked with my parents neither one of them shares a single Mb with her, neither do my paternal grandparents. She is almost fully of Northern European descent, whereas I'm 85%+ of Southern European ancestry, that has to be one of the biggest IBS segments ever. What sucks is she seemed very happy to connect with me. How do I break the news to her now that we aren't really related.

J Man
12-05-2014, 02:57 PM
Check this out:

Today somebody added me, with whom I share 31 Mb, yet when I checked with my parents neither one of them shares a single Mb with her, neither do my paternal grandparents. She is almost fully of Northern European descent, whereas I'm 85%+ of Southern European ancestry, that has to be one of the biggest IBS segments ever. What sucks is she seemed very happy to connect with me. How do I break the news to her now that we aren't really related.

Mb? Is that a different way of measuring compared to cM?

ArmandoR1b
12-05-2014, 04:01 PM
Check this out:

Today somebody added me, with whom I share 31 Mb, yet when I checked with my parents neither one of them shares a single Mb with her, neither do my paternal grandparents. She is almost fully of Northern European descent, whereas I'm 85%+ of Southern European ancestry, that has to be one of the biggest IBS segments ever. What sucks is she seemed very happy to connect with me. How do I break the news to her now that we aren't really related.

Using Mb is supposedly unreliable and it is probably why that happened. http://dna-footprints.com/677/ancestrydna-cm-mb-disparity/

I wonder what the largest segment is in cMs. Have you uploaded your file and the files of your parents and paternal grandparents to Gedmatch? If not you should do that and then have your "match" do the same with her's and her parents, if she has their files, and then do one-to-one with 400 SNPs and 4 cM.

jeanL
12-05-2014, 09:00 PM
Using Mb is supposedly unreliable and it is probably why that happened. http://dna-footprints.com/677/ancestrydna-cm-mb-disparity/

I wonder what the largest segment is in cMs. Have you uploaded your file and the files of your parents and paternal grandparents to Gedmatch? If not you should do that and then have your "match" do the same with her's and her parents, if she has their files, and then do one-to-one with 400 SNPs and 4 cM.

I just checked she shares 12.1 cM with me, but nothing with either one of my parents!

SwampThing27
12-05-2014, 09:24 PM
I thought this thread was going to be about problems in the restroom.

Gray Fox
12-05-2014, 09:38 PM
I thought this thread was going to be about problems in the restroom.

I wasn't too sure myself :heh:

ArmandoR1b
12-06-2014, 01:02 PM
I just checked she shares 12.1 cM with me, but nothing with either one of my parents!

That's not too surprising. As I had posted in the 2nd post of this thread - IBS can be as large as 12.9 cM See http://www.jogg.info/72/files/Turner.htm

At the bottom of that page she has -

For instance, one child had 17/70 matches not found in either parent. The lengths of these were:

7.70
7.72
7.75
7.76
7.80
7.82
7.94
7.97
7.97
8.03
8.12
8.28
8.81
9.45
9.70
11.93
12.90

Jesse1961
06-22-2016, 11:34 AM
As Armando alludes to, it's really a question of probabilities. From comparing multiple kits of my relatives I can see there are some surprisingly short IBD segments (around 3 cM) but the odds of that in two random people must be rather low...

If some persons triangulate at a certain 3cM segment that is considered IBD segment, how many generations ago did their common ancestor lived?

AJL
06-22-2016, 03:05 PM
If some persons triangulate at a certain 3cM segment that is considered IBD segment, how many generations ago did their common ancestor lived?

Impossible to say.

loisrp
06-22-2016, 05:48 PM
Ancestry.com does its phasing, and applies the TIMBER algorithm. If a match appears, after having been filtered through this Ancestry processing, how high a chance is there of the match being IBS? (I realize no one may have the exact answer.)

How about for a match on FamilyTreeDNA being IBS? The cM values are always lower than I see on Gedmatch, so I assume some filtering is being applied.

There has to be some sort of clustering algorithm that could be applied, although I'm unsure how to do it. For example, I might not know who our common ancestors are for my numerous small matches in Appalachia, but there are so many on different chromosomes that I can't simply discard them as being IBS.

crossover
06-23-2016, 01:09 AM
That's not too surprising. As I had posted in the 2nd post of this thread - IBS can be as large as 12.9 cM See http://www.jogg.info/72/files/Turner.htm

At the bottom of that page she has -

For instance, one child had 17/70 matches not found in either parent. The lengths of these were:

7.70
7.72
7.75
7.76
7.80
7.82
7.94
7.97
7.97
8.03
8.12
8.28
8.81
9.45
9.70
11.93
12.90

is it possible for an ibs segment to be 15cm +. unfortunately, all my great grandparents passed away before decent dna analysis technology came to be, so i can't phase the dna i got from my surviving grandparents

loisrp
06-23-2016, 01:14 AM
15 cM+ is pushing it, but I don't have enough empirical evidence to say. It could also be two segments that appear next to each other without a visible break, particularly in an endogamous community. Why do you think the person is not related?