Search for content in message boards

Haplogroup prediction from raw data?

Haplogroup prediction from raw data?

Posted: 21 Apr 2013 2:49PM GMT
Classification: Query
Does anyone know of a program that will give me my haplogroup prediction from my ancestry dna raw data (autosomal)? I would imagine there is a way to do this through snp groupings but perhaps I'm mistaken. I was hoping promethease would've done that for me but haplogroup prediction wasn't in the file package from my download.

Re: Haplogroup prediction from raw data?

Posted: 24 Apr 2013 10:39PM GMT
Classification: Query
Edited: 24 Apr 2013 10:40PM GMT
The issue is that the SNPs ancestry selected were done to fit their product - cousin matching and ethnicity. It just so happens that some of them are related to haplogroups as well as medical info, and that is what Promethease looks at. The more obvious question is I can't tell if you are male or female, and if you are female you don't have a Y chromosome in the first place. Female line haplogroups are tested in mitochondrial DNA or mtDNA, and the autosomal test does not look at that.

If you are male, just to make sure go to the "Topics" section of the Promethease report and look for Haplogroups.

David Mc

Re: Haplogroup prediction from raw data?

Posted: 24 Apr 2013 11:38PM GMT
Classification: Query
I'm male and when I ran the promethease report on my AncestryDNA data it said I was in the I haplogroup. From a yDNA test I had done several years ago I'm actually I1 so it can give you an idea of your general haplogroup, it just can't tell you the specific subgroup you are in.



Re: Haplogroup prediction from raw data?

Posted: 25 Apr 2013 12:23AM GMT
Classification: Query
How do you actually interpret what Promethease is reporting to come up with a haplogroup? My report shows 4 Haplogroup-related entries in the topics section:
1) Haplogroup I (Y-DNA) - contains a single entry that says I'm M258-
2) Haplogroup R1b1b2g (Y-DNA) - contains an entry at says it is M222, but doesn't say if my genotype (A;A) is positive or negative, and the summary section when I pull up the page on SNPedia says "relevant to haplogroup R1b1b2e" (which one is it, R1b1b2e or R1b1b2g?). There are also two other entries in this section, but they just say common for Caucasians. I guess that might mean I'm negative for those, but it's not clear what that means.
3) Haplogroup Y (Y-DNA) - contains many entries, but most of them say nothing until you click "more", and even then just say "Haplogroup: ?", which doesn't tell me anything. When I pull up the pages on SNPedia the just say they distinguish haplogroups, but doesn't say in what way. A few of the entries say "[haplogroup: R1]", "[haplogroup: P]", "[haplogroup: F-R]", "[haplogroup: R]", "[haplogroup: R1]", "[haplogroup: IJ]", "common on affy axiom data". There is also a repeat of the M222 SNP (and I assume of the two blank ones as well) and another that says I'm M72-.
4) Haplogroups - appears to contain a repeat of everyting from the Haplogroup Y (Y-DNA) section.

There doesn't appear to be very much useful information entered into SNPedia for these, and I can't make heads or tails of it for trying to guess my own haplogroup.

Re: Haplogroup prediction from raw data?

Posted: 25 Apr 2013 12:50AM GMT
Classification: Query
Edited: 25 Apr 2013 12:50AM GMT
Jeff/Doug,

Basically, if ancestry happened to test a SNP that was an indicator of a deep haplogroup and you are positive for that, then lucky you. Otherwise, you might just get something very common like R1b, or none at all.

The best thing to do is look at the ISOGG information because the haplogroups are always changing and Promethease is using an out of date tree. The following link is to a list of the SNPs related to haplogroups and what they indicate. Promethease will tell you the SNP, and then you can look it up at ISOGG.

http://www.isogg.org/tree/ISOGG_YDNA_SNP_Index.html

You can then look at the the haplotree to see where your haplogroup fits in.

http://www.isogg.org/tree

You will also be a member of all the groups above the one you tested positive for.

David Mc

Re: Haplogroup prediction from raw data?

Posted: 25 Apr 2013 12:55AM GMT
Classification: Query
This post was deleted by a board administrator on 28 Apr 2013 6:45PM GMT

Re: Haplogroup prediction from raw data?

Posted: 25 Apr 2013 1:50AM GMT
Classification: Query
David,
Thanks for the great response.
Indeed, I am a male, so the Y chromosome data is there.

Forgive my follow up but I'm a bit of a newb with autosomal data and it's been quite some time since my last genetics class.
When trying to determine my own estimated haplogroup(or when an analyst, or program does it), what exactly am I pooling together to make that determination? For example, of the 125 SNPs listed, there are a handful of haplogroups classified. The ones shown on the Promethease report range from R1, P, F-R, R, I, and IJ.
I happened to look up all of the other SNPs that weren't classified, and they're all over the board too.
Is it as simple as looking at the greatest number of common haplogroups to make my own predicted haplogroup classification, or is that not really possible without knowing the STRs and/or frequency?

Thanks in Advance,
Mike

Re: Haplogroup prediction from raw data?

Posted: 25 Apr 2013 2:53AM GMT
Classification: Query
Edited: 25 Apr 2013 2:55AM GMT
Mike,

I'm used to approaching this from the other direction (Y-DNA) so I'm not sure if I can answer. I will say that as far as I know it is not a numbers game - your haplogroup is determined by specific values for specific SNP(s), you either are or you aren't. I know my haplogroup from doing an SNP test at FTDNA, so when I first looked at my Promethease report I just saw the R1b... which is correct although it uses an older designation. However, I just went back and saw the I and Y (which doesn't exist, maybe they meant Y chromosome?) so I'm not sure what they are trying to tell you. The author of Promethease said in another thread that the focus was on medical and not haplogroups, so there could be some errors in how it is interpreted.

Typically you take a Y-DNA test which gives a series of markers called STRs. Your specific sequence of those markers is called your haplotype. Based on that, they can predict your haplogroup because other people with the same or similar haplotype have tested positive for some specific SNP(s). If you want to carry it further down the tree, you can get tested for whatever SNP is appropriate for your grouping.

I think the problem with ancestry DNA (and thus Promethease) is that the SNP picture as far as haplogroups is incomplete. For example, they may have tested 2 important ones but not a third, so you'll get incomplete or conflicting results. There will also be haplogroups that won't be represented at all. And as mentioned, it could simply be mistakes in their database.

David Mc

Re: Haplogroup prediction from raw data?

Posted: 25 Apr 2013 5:42AM GMT
Classification: Query
Promethease and the link you gave in you last message both show rs20321 as the SNP for M222. The web site says the mutation there is G -> A, and my genotype for that position is A;A according to my Promethease report, so it looks like I'm positive for M222. Based on the website you gave that would make my haplogroup and subclade definitely R1b1a2a1a2c1a1a1? At least that's the name that other web site gives it. Is "R1b1c7" an older name for the same haplogroup? The G;G entry for that same SNP in SNPedia gives that "common on affy axiom data" message that I was getting on several of SNP's, so that would confirm my theory that those messages meant they were negative for whatever mutation they were testing. Someone should go through the list you linked to and enter the correct and current full haplogroup subclade names and the names of the specific mutations and alternate/previous names stating whether it was positive or negative for that mutation to make it clear to the average person who wouldn't necessarily know that "common on affy axiom data" means negative for that marker.

Re: Haplogroup prediction from raw data?

Posted: 25 Apr 2013 11:02AM GMT
Classification: Query
@Jeff - Agreed. Determining if the SNP is in the derived state, or not seems confusing. I am going to have re-confirm my manually looked-up classifications as to whether they are derived, or non-mutated.
Why wouldn't they just say, "non-mutated" instead of "common on affy axiom data?"
Those R1b1 classifications may be the same depending on the source. I've gotten thrown off a few times by seeing updated classifications, and subclades.
per page

Find a board about a specific topic

  • Visit our other sites:

© 1997-2014 Ancestry.com | Corporate Information | New Privacy | New Terms and Conditions