PDA

View Full Version : Question regarding the transmission of dna from parent to child



leonardo
11-01-2014, 12:20 PM
I hope somebody within this community can answer this question. Please feel free to correct any misconceptions. As I understand things, parents pass autosomal dna on to a child, each at about 50%. This autosomal dna is passed on randomly, in the sense that two siblings could have a slightly different dna structure. Here is my question, can a parent who has a small piece of autosomal dna pass on a larger share to a child? More specifically, if a parent has a small match with another person (say 4cM), could their child end up sharing 30cM - including a single match on a chromosome of 7.5?

dp
11-01-2014, 04:46 PM
4 cM is small enough that that could be an IBS segment. A minimum 10 cM (IBD) block is used for (most) genetic genealogical relationships.
If a parent has a common match, as does their child, I would expect the degree of match to be less. I suggest that you (or they) run the parents related util for the child (assuming the other parent cant be tested) on GEDmatch. If it picks up a relationship both of the child's parents are related --in my experience to a larger degree than indicated. Thus, the larger descent could be coming from the missing parent.
dp :-)
In other words because my parents are related I match a few people (better) than my mother. In some cases I have matches that she doesn't have, nor my father's brother.

leonardo
11-01-2014, 05:28 PM
Thanks David. This is from Gedmatch. I have read their parameters differ from FTDNA or 23andMe. My mother and I tested with FTDNA. The person in question with 23andMe. My mother has a match on chromosome 20 with this individual at 4.5 cM and 392 SNP, likely a IBS match I would say. I have matches on 6 chromosomes (but not 20 like her) with this individual. My largest is 7.5cM and 1868 SNP. The next largest is 5.2. All told, they total 30cM. I wonder if I could share more with this individual who is my mother's age, than she does? If I share with this man through my mother, could she have passed on to me more of a match than what she has? By the way, the reason I dropped the parameters down is because my father and mother have common ancestries (they are not related though - at least according to Gedmatch). So, when I use the defaults settings I am always worried that it may appear to be a match just for me, and not my mother - but, then she often has matches at 5cM or so and maybe 500 or 500 SNPs. I then have this sense of doubt, 'does this person share with my mother as well, or is this a match through my father?'

ilmari
11-01-2014, 08:22 PM
No Call results in a parent's test can also break a segment into smaller pieces that fall outside of the parameters. I have some like that with my mom, which is easy to see if you look through your raw data, but there are other, more common, reasons that people have more than a parent might with another individual, and that is pedigree collapse and interrelated couplings.

leonardo
11-01-2014, 08:43 PM
Thanks ilmari. Now, would something like this be common among ancestry that is likely from the Mennonite community?

ilmari
11-03-2014, 08:47 AM
Well, I would be prone to say yes, but that is just off of the top of my hat. I am Finnish, and pedigree collapse is a common occurrence in Finns. I've seen the same in other isolated groups, so if they were strict adherents to their community, I would say very likely, and it could just come from one grandparent.

Tolan
11-03-2014, 10:01 AM
I have also a question about this:
On one pair of base, i think the DNA testing companies can not know what comes from the father or the mother. How can they know the whole segments from one or the other parents?
To make me understand, an example below:
For example, I have 3 pairs: AT TT GA, ATG comes from my father, and TTA from my mother. There are four possible combinations, two are good, two are false.
Do I also have matches for falses combinations?
But maybe I'm wrong in my deductions!

2861

dp
11-04-2014, 09:15 PM
Tolan,
You only have two copies of each autosomal chromosome (assuming we're not dealing with chimerial DNA, or allosome [like XXY]). If you have a full match for over a certain amount of bases between your chromosome pairs then your parents are related. In other words I dont understand what you mean by ATG and TTA.
dp :-)

vettor
11-08-2014, 09:48 PM
I doubt we can ever tell how much of each parents genes went into each child.

as for me

my oldest son, has 47.6% of my genes and 52.4% of his mothers
next son , has 50.9% of my genes and 49.1% of his mothers
others not tested

We cannot tell really anything on the exact composition within any genetic parents our cousins

Dimanto
11-14-2014, 12:36 AM
4 cM is small enough that that could be an IBS segment. A minimum 10 cM (IBD) block is used for (most) genetic genealogical relationships.
If a parent has a common match, as does their child, I would expect the degree of match to be less. I suggest that you (or they) run the parents related util for the child (assuming the other parent cant be tested) on GEDmatch. If it picks up a relationship both of the child's parents are related --in my experience to a larger degree than indicated. Thus, the larger descent could be coming from the missing parent.
dp :-)
In other words because my parents are related I match a few people (better) than my mother. In some cases I have matches that she doesn't have, nor my father's brother.

No I don't think so because I have several matches that I'm matching and sometimes almost double than my mum while my parents are both from different countries separated more than 1900 km, who hadn't have recent gene flow. However in some cases some matches also had a few % or less Italian & S European and matched on 2 different segments from which one overlapped with my mum. In some cases the overlap was there but a little bit longer than the one shared with my mum on one specific chromosome.

I also found several matches on my mum's DNA relatives which I don't even have who're in her top 10 matches. My top match is on her third page or so... Can anyone explain how that can happen ?

dp
11-14-2014, 06:25 PM
No I don't think so because I have several matches that I'm matching and sometimes almost double than my mum while my parents are both from different countries separated more than 1900 km, who hadn't have recent gene flow.
Dimanto,
Here's an idea. Have GEDmatch make phased kits from your's and your mom's kits. Leave the new kits set as research --so they don't get into others one-to-many searches. Run the one to many for the new kit and see how that compares to your original kit. Maybe some of these "matching" segments are IBS, and perhaps this will drop out false matches.
-dp

Dimanto
11-16-2014, 04:58 AM
Dimanto,
Here's an idea. Have GEDmatch make phased kits from your's and your mom's kits. Leave the new kits set as research --so they don't get into others one-to-many searches. Run the one to many for the new kit and see how that compares to your original kit. Maybe some of these "matching" segments are IBS, and perhaps this will drop out false matches.
-dp

I understand. Thank you very much for your suggestion; I'll surely follow your instructions !

Tolan
11-18-2014, 04:21 PM
I have also a question about this:
On one pair of base, i think the DNA testing companies can not know what comes from the father or the mother. How can they know the whole segments from one or the other parents?
To make me understand, an example below:
For example, I have 3 pairs: AT TT GA, ATG comes from my father, and TTA from my mother. There are four possible combinations, two are good, two are false.
Do I also have matches for falses combinations?
But maybe I'm wrong in my deductions!

2861

Someone gave this link in another thread, which explains probably better than me, why a child may have a match that parents don't have.
This is a "IBS" or a false match

See: http://www.jogg.info/72/files/Turner.htm

dp
11-18-2014, 09:38 PM
Someone gave this link in another thread, which explains probably better than me, why a child may have a match that parents don't have.
This is a "IBS" or a false match

See: http://www.jogg.info/72/files/Turner.htm

Very thorough. Thanks for the link.
dp :-)

Dimanto
11-20-2014, 03:48 PM
Thanks but no thanks

Pillar_of_fire
11-21-2014, 06:48 PM
Hi, Not exactly for this thread, but did not want to start a new one...

How to calculate percentage of shared cM between siblings?

Applying the shared cM to 6770 I usually receive less than 50%


brother - sister - 2594.04 176.65 38.32%

sister - sister 2773 267 40.96%

brother - sister 2469.44 183.09 36.48%

brother - sister2273 205 33.57%

Should I do some adjustment because they share the two parents?

dp
11-26-2014, 03:31 PM
Hi, Not exactly for this thread, but did not want to start a new one...

How to calculate percentage of shared cM between siblings?

Applying the shared cM to 6770 I usually receive less than 50%


brother - sister - 2594.04 176.65 38.32%

sister - sister 2773 267 40.96%

brother - sister 2469.44 183.09 36.48%

brother - sister2273 205 33.57%

Should I do some adjustment because they share the two parents?

Dear POF,
I have recently come across a similar instance on GEDmatch comparisons:
Parent -child: 3587 (consistently)
sister-sister: 2652
uncle - nephew: 1893
aunt - nephew: 1744
second cousins: 140-175

I was under the impression that siblings should be at least half-match over the whole autosomal genome. That isn't the case with the sister-sister pair (based on GEDmatch comparison). Where did the (about) 1000 cM difference come about from?

David Powell
dp :-)