Family Finder SNPs

[madman]

you have C, C

son has C, T

other parent has T, ?

FTDNA could put in all the major alleles in the raw data and then we could extract those with the minor alleles.
I still dont understand why I have a CC at certain SNPs and my son has say TC.

[PeterLarsen]

How then did I end up with the CC in the first instance, one each from my father and mother? If I got a FF test for my son it would be a lot of work separating my SNPs from his mothers.

You have CC because you inherited C from your father and C from your mother. From your data you can only know 2 of your parents 4 allele, and the other 2 is unknown unless you test your parents or a sibling or other relative to find it out. That is phasing, and can be extended into more useful things. It is starting to get useful when you have results from more close relatives.

[1798]

Thanks to all your replies.I would like to get a FF test for my son. I am not that old but I am the oldest member of my family line left in Ireland.

[Ann Turner]

I got an email from FTDNA saying that all the 69,000 plus SNPs in my FF test are specific to me.

I guess that hinges on how they interpreted the word specific. You do have your own results. However, everybody has results for the same SNPs, and many people will share one or both of your alleles. This is the sense that I meant: there are no SNPs where you're the only person with a particular result.

[1798]

I guess that hinges on how they interpreted the word specific. You do have your own results. However, everybody has results for the same SNPs, and many people will share one or both of your alleles. This is the sense that I meant: there are no SNPs where you're the only person with a particular result.

My son doesn't have the exact same alleles that I have for some of my FF SNPs and I don't have any exact matches at FF, so that means that everyone does not share the same alleles for the SNPs tested.
Here are a few examples of my matches;
match#1 Ch. 12, 9.42, 2,300 SNPs
match#2 Ch. 16, 10.15, 1,499 SNPs
match#3 Ch. 7, 8.3 , 2,100 SNPs
match#4 Ch. 1, 8.09, 1176 SNPs.
Those are all a long way from 690,000.

[Ann Turner]

My son doesn't have the exact same alleles that I have for some of my FF SNPs and I don't have any exact matches at FF, so that means that everyone does not share the same alleles for the SNPs tested.
Here are a few examples of my matches;
match#1 Ch. 12, 9.42, 2,300 SNPs
match#2 Ch. 16, 10.15, 1,499 SNPs
match#3 Ch. 7, 8.3 , 2,100 SNPs
match#4 Ch. 1, 8.09, 1176 SNPs.
Those are all a long way from 690,000.

Your son has two results (alleles) for every SNP. One of them came from you, and the other from his mother. There's no way to tell from his results alone which allele came from which parent. If you have data for all three people -- father/mother/child -- you can often (but not always) deduce which allele came from which parent.

FF looks for long consecutive runs of SNPs where at least one of your alleles matches at least one of the other party's alleles. You will match the other party for hundreds of thousands of alleles, but they won't be consecutive.

It's a bit like asking two people to toss a coin 1000 times. They might have matching results on toss #3,#9,#13,#15,#16,#21 etc. etc. just by chance. But if you see the two people match exactly for a section of 100 SNPs in a row, you'd be suspicious that they were actually doing the work of tossing the coin independently. It would be more likely that one person got tired and copied the other person's results for a while. Your FF report shows sections where you and your match have copied the same portion of DNA from a common ancestor.

[1798]

I picked out three SNPs at random from my son's Geno2 autosomal dna results and checked them against mine.
Son Me
rs2837565 TC TT
rs2838416 TC CC
rs2839050 AG AA
I think that you could say for sure in each case that he got one of these alleles from me and the other from his mother.
Do you think that is correct? It would take me a lot of time but I am sure I could extract all his SNPs relative to mine using both tests.

[Ann Turner]

I picked out three SNPs at random from my son's Geno2 autosomal dna results and checked them against mine.
Son Me
rs2837565 TC TT
rs2838416 TC CC
rs2839050 AG AA
I think that you could say for sure in each case that he got one of these alleles from me and the other from his mother.
Do you think that is correct? It would take me a lot of time but I am sure I could extract all his SNPs relative to mine using both tests.

Yes, that's correct. David Pike has written a utility to automate the process:

http://www.math.mun.ca/~dapike/FF23utils/trio-phase.php

[1798]

Yes, that's correct. David Pike has written a utility to automate the process:

http://www.math.mun.ca/~dapike/FF23utils/trio-phase.php

Thanks for the link.
I must have autosomal snp variants that are not found outside of Ireland. Would you agree with that statement?

[Ann Turner]

Thanks for the link.
I must have autosomal snp variants that are not found outside of Ireland. Would you agree with that statement?

Sure -- if they originated with an ancestor whose descendants all stayed in Ireland. But they won't be found on a test like Family Finder. You would need to do whole genome sequencing to discover them.

[1798]

Sure -- if they originated with an ancestor whose descendants all stayed in Ireland. But they won't be found on a test like Family Finder. You would need to do whole genome sequencing to discover them.

Why not?!!!

[Ann Turner]

Why not?!!!

There are only ~700,000 SNPs on the chip. They were selected because the minor allele is somewhat common in the HapMap or 1000 Genomes Project databases (at least 5%). Those databases don't include samples collected in Ireland. It wouldn't be efficient for a mass-produced chip to include variants that are seldom found.

The fact that the minor allele is somewhat common means that the mutation had to happen long enough ago that it has had a chance to rise to a certain frequency. This may be tens of thousands of years ago, and in some cases, hundreds of thousands of years ago. Some variants found in mankind today existed before anatomically modern Homo sapiens was around. The consequence is that those variants have had a chance to migrate around the world. Thus there is very little specificity. There are virtually no SNPs where one allele is found in 100% of one population and 0% of all other populations. The best we can do is look at differences in frequency.

[1798]

There are only ~700,000 SNPs on the chip. They were selected because the minor allele is somewhat common in the HapMap or 1000 Genomes Project databases (at least 5%). Those databases don't include samples collected in Ireland. It wouldn't be efficient for a mass-produced chip to include variants that are seldom found.

The fact that the minor allele is somewhat common means that the mutation had to happen long enough ago that it has had a chance to rise to a certain frequency. This may be tens of thousands of years ago, and in some cases, hundreds of thousands of years ago. Some variants found in mankind today existed before anatomically modern Homo sapiens was around. The consequence is that those variants have had a chance to migrate around the world. Thus there is very little specificity. There are virtually no SNPs where one allele is found in 100% of one population and 0% of all other populations. The best we can do is look at differences in frequency.

So you are saying that the alleles I have for all 900,000 SNPs are same for everyone who does the FF test.

[Ann Turner]

So you are saying that the alleles I have for all 900,000 SNPs are same for everyone who does the FF test.

No, but everyone has the same list of SNPs. Virtually every SNP on the chip has just two possible alleles. Everyone will have one or the other allele, perhaps two copies of the same allele (e.g. AA or GG) or one copy of each(AG).

[1798]

No, but everyone has the same list of SNPs. Virtually every SNP on the chip has just two possible alleles. Everyone will have one or the other allele, perhaps two copies of the same allele (e.g. AA or GG) or one copy of each(AG).

I understand and thanks. I would like FTDNA to put a column in the raw data with all the major alleles for each SNP. There are some autosomal SNPs that are passed on then unchanged for thousands of years and others that recombine and are recent events.