Which Y and Mtdna tests?

[Arimas]

I have decided to take both Y and Mtdna tests to know who were my further ancestors about 20 generations ago. I know my male family tree and my dad was from the 20th generation from my furthest known ancestor according to our oral tradition. I also know the supposed ethnic origin of that ancestor.

I was thinking I need less markers than Americans or Europeans looking for close cousins with their family names. The equivalent of family names are clans and tribes in our culture with known family trees so one can retrace their ancestry up to the known ancestor...if the is no reason to suspect a "natural" or an adopted child on the way.

On my Mtdna, I would maybe need more details as I know less on my great...great grandmas. Also, some traditions says that the female ancestors of many Somalis were of Cu****ic Oromos who married male Arab ancestors. Can the MtDna say if my further known female ancestor was an Oromo or not?

Would the basic Y and Mtdna be enough to know for sure the haplogroups? I read on this board that a 12 marker test (basic Y) is only to predict the haplogroup so I am wondering which one can say it without doubt. Close matches are not what I need to know and anyway they wouldn't be many Somalis in the databases.



Arimas

[cacio]

Arimas:

the basic Y and mtdna tests as offered by the Genographic project may or may not be enough to determine haplogroups, it depends (for instance, it wasn't enough for my Y haplogroup, since it is not very common in Europe).

As for matches, if you test a cousin, or a person documented to be on the same male line, and you have the same 12 STR markers, then you can pretty much establish relation, as you say in your post. However, if you test a random person and he turns out to have the same 12 markers, then you cannot really say for sure that there is a close relation (ie within 200 years or so) - this may have happened randomly.

So my suggestion is to start with a basic 12 STR markers test with the genographic project. If you want to doublecheck your haplogroup, you can take a test with Ethnoancestry, which will directly verify your haplogroup (and subhaplogroup), but without telling you your STRs. (As I guess is clear, the mutations defining haplogroups are not in the STRs).

As for the mtdna, the genographic project/FTDNA tests for HVR1 (and for additional money the HVR2), which is usually enough to say the haplogroup. If you can spend a little more money, argusbio.com is offering a cheap sequence of the whole mtdna for $350 (the whole 16600 bases, not just the last 600 as the HVR1 test). So that gives much more information, although it is more expensive.

At this point, you will not be able to determine whether the mtdna is Oromo or not- as said before, mtdna's tend to be spread all over Africa (although some are present predominantly in Ethiopia), and there aren't yet that many samples from Africa to compare. The whole mtdna (given its size) may one day give enough information to pinpoint the geographic origin, but only when a much more broad database is available.

cacio

[tomcat]

FTDNA hosts an African Project - got to FTDNA homepage, in upper right search for project name that includes 'Africa' follow the links to see if this project is appropriate for you. They offer a group rate for y-DNA (12 marker) and mt-DNA (HVR 1) combined for $199.

FTDNA also offers offer deeper testing to confirm haplogroups etc. There is a cost saving to be realized by consistently testing with one lab.

[vineviz]

Can the MtDna say if my further known female ancestor was an Oromo or not?

Probably not. HVR-1 results alone will be unlikely to seperate an Oromo descendant from (say) a Yemeni or Amahara descendant. In fact, I'm not sure that any level of mtDNA test would conclude Oromo ancestry with any certainty. The full-sequence mtDNA would probably be required, and even then you'd have to hope that the results matched an academic paper (e.g. http://www.journals.uchicago.edu/AJH.../41578.web.pdf )

Would the basic Y and Mtdna be enough to know for sure the haplogroups? I read on this board that a 12 marker test (basic Y) is only to predict the haplogroup so I am wondering which one can say it without doubt. Close matches are not what I need to know and anyway they wouldn't be many Somalis in the databases.

The 12 marker test alone will be insufficient to establish your subclade. And some haplogroups of interest to Somalis (like E-V6) are not detectable by any commerical test at all. Even if you are not trying to match "cousins", the 25 and 37 marker panels include some very useful and diagnostic STRs for haplgroup sub-clade.

http://www.journals.uchicago.edu/AJHG/journal/issues/v74n5/40866/40866.web.pdf

[Arimas]

OK, it make sense to start with a 12 markers.

I don't have a particular reason to think that my female ancestors were Oromo but usually it is understood that the local women were Ethiopian (like of my mom's ancestor is called the son of the Habash woman...I think it make more sense that those called Habash (Abyssinian) were actually Oromo since the real Habash live in Northern Ethiopia.

What is are the STRs? And HVR1 and HVR2?

What can I learn from the 16660 bases?


Thanks Cacio,


Arimas

So my suggestion is to start with a basic 12 STR markers test with the genographic project. If you want to doublecheck your haplogroup, you can take a test with Ethnoancestry, which will directly verify your haplogroup (and subhaplogroup), but without telling you your STRs. (As I guess is clear, the mutations defining haplogroups are not in the STRs).

As for the mtdna, the genographic project/FTDNA tests for HVR1 (and for additional money the HVR2), which is usually enough to say the haplogroup. If you can spend a little more money, argusbio.com is offering a cheap sequence of the whole mtdna for $350 (the whole 16600 bases, not just the last 600 as the HVR1 test). So that gives much more information, although it is more expensive.
cacio

[Arimas]

FTDNA hosts an African Project - got to FTDNA homepage, in upper right search for project name that includes 'Africa' follow the links to see if this project is appropriate for you. They offer a group rate for y-DNA (12 marker) and mt-DNA (HVR 1) combined for $199.

FTDNA also offers offer deeper testing to confirm haplogroups etc. There is a cost saving to be realized by consistently testing with one lab.

Thanks Tomcat,
199$ for both tests is good...that's about the amount I wanted to take from my hobby budget. I think no amount would have been big if I was looking for lost family but in my case, many of my distant cousins are around and they're poor...makes me sometimes wonder if they don't deserve my money instead of a chimeric ancestry.
But I also value learning and education so maybe I should think I am investing in my education and buy the most refined tests.

FTDNA would be my first choice as they host this forum where I am learning so much.


Arimas

[Arimas]

Probably not. HVR-1 results alone will be unlikely to seperate an Oromo descendant from (say) a Yemeni or Amahara descendant.

It is too bad...

The 12 marker test alone will be insufficient to establish your subclade. And some haplogroups of interest to Somalis (like E-V6) are not detectable by any commerical test at all. Even if you are not trying to match "cousins", the 25 and 37 marker panels include some very useful and diagnostic STRs for haplgroup sub-clade.

I hope I'll be from an easy haplogroup if not, I'lll go to the 37 markers.


Thanks Vineviz, for the links. I appreciate your help.


Arimas

[vineviz]

I hope I'll be from an easy haplogroup if not, I'lll go to the 37 markers.

Also, for your purposes, after you get your 12 marker results back there are probably a couple of "a la carte" markers that will help you refine your haplogroup prediction without paying for the full 37 marker upgrade. Or even a deep-SNP test if your broad haplogroup is well-predicted by 12 markers might be a good investment.

[efgen]

Something else to consider is that testing through the Genographic Project practically guarantees an accurate Y-DNA haplogroup assignment. If they can unambiguously predict your haplogroup, then they will do so. However, if they can't, then they'll do a free SNP test to find out your haplogroup. This differs from FTDNA in that if FTDNA can't predict your haplogroup, then you have to pay separately for the SNP test.

So if a definite Y-DNA haplogroup is most important to you, then I would recommend starting with the Genographic Project. Once you get your results back (12 markers + haplogroup), you can transfer your results to FTDNA for free -- then order additional tests, such as more markers, deep-clade and mtDNA tests. (Note: FTDNA's lab does the testing for the Genographic Project, so you're not actually transferring your DNA to another lab -- rather, you're just transferring your results from the GP database to the FTDNA database and getting an account on FTDNA's website).

Also, I don't think I saw anyone mention that FTDNA now always determines mtDNA haplogroups by running a panel of SNP tests (the list of SNPs is not disclosed). Yes, this is somewhat new (past year or so?), so those who tested before this change may only have a predicted haplogroup.

And finally, none of these statements refers to deep-SNP (subclade) testing. Deep-SNP tests are always optional and at an additional cost.

[lgmayka]

I wholeheartedly agree with efgen that for anyone with paternal ancestry other than Western European, the Genographic Project is an excellent choice, because it guarantees determination of your haplogroup--via an SNP test, if necessary:

https://www3.nationalgeographic.com/genographic/participate.html

Almost all of my (rural southern) Polish relatives so far have had haplogroup ambiguity, in which the initial 12 markers of the haplotype are insufficient to confidently predict the haplogroup.

Once you have your Genographic results, you can transfer them (at no charge) into an FTDNA account. You can then compare your results against FTDNA's customer database, order more tests, join DNA projects of interest to you, etc.

For mtDNA, I must point out that ArgusBio offers HVR1+HVR2 testing for only $75:

http://www.argusbio.com/store/product_info.php?products_id=28

FTDNA charges about $160 for both HVR1+HVR2.

The primary disadvantage of ordering from ArgusBio is that you will be unable to compare the results against FTDNA's large customer database. You will be able to enter ArgusBio results into the public Ysearch database, but that one is much smaller than FTDNA's private one.

[Arimas]

Oh! My understanding was that the Genographic project and FTDNA were with the same company. Thank you for the clarification and all your tips Efgen.

This is why I like this forum: I learn from every post.


Arimas

[Arimas]

Also, for your purposes, after you get your 12 marker results back there are probably a couple of "a la carte" markers that will help you refine your haplogroup prediction without paying for the full 37 marker upgrade. Or even a deep-SNP test if your broad haplogroup is well-predicted by 12 markers might be a good investment.

Thanks again Vineviz...I am getting closer I think.

[Arimas]

thanks lgmayka for the feedback and tips.

I am getting addicted to the website now

[cacio]

Arimas:

regarding your questions.

On the mtdna: the mtdna is a string of around 16600 molecules, represented by the letters A,C,G,T. Haplogroups are determined by the particular sequence of these 16600 letters. For instance, say, I have the letter C at position 16189, while most other people have, say, T. The basic test, HVR1=hypervariable region 1, looks only at 600 locations, number 16000 to 16600 roughly. HVR2 looks at 1 to 600. These are the most informative parts, but that leaves out 601-15999, which contain lots of other informative locations. So as usual, it's a matter of price/information ratio. You get a lot of information with HVR1, but you get much more by testing the whole thing.

Efgen: by the way, I've heard this repeated many times that FTDNA tests for specific locations outside HVR to determine haplogroups, but I have seen too many recent mis-assignments for this to be true. May be they do this for European haplogroups like H or T, but not in general.

As for the Y-chromosome test, I think there is some confusion between haplogroups or haplotypes.
Haplogroup= very ancient lineage (like E3b etc.), ie we're talking about many thousand year old lineages. This is determined by specific letters at certain location of the Y chromosome. These locations are called SNP. So an SNP test would lookfor a certain letter at a specific location of the Y chromosome, defining E3b (or whatever other haplogroup).

Haplotype= set of STR=short tandem repeat. Somewhere in the Y-chromosome there are sequences of letters repeated many times (like GATAGATAGATAGATA). The basic test will tell you 12 STR, that is the number of repeats in 12 particular places. Because these numbers of repetitions vary very frequently, they are useful to determine close ancestry, that is, whether the two people are on the same patrilinear line within a couple hundred years or so. However, STR don't directly determine the haplogroup (E3b etc.). Haplogroups can be predicted by the STR simply by looking at big databases. For instance, if on the database there is another person who has the same 12 STR markers and has been determined to belong to a certain haplogroup, then one would assume you are related and you belong to the same haplogroup. But if there is nobody close to you, then the STR won't tell you the haplogroup. And it is always possible that there's another person who by chance has the same 12 STRs but belongs to a different haplogroups.

The genographic project is tested by FTDNA. However, what efgen was saying is the following. If you buy the STR directly from FTDNA, they will only test the 12 location. If they cannot predict the haplogroup, then that's it. If you go through the genographic project, they will first test the 12 STR markers. If they can predict, fine. Otherwise, they will try doing some SNP test to determine also the haplogroup (eg E3b etc.).

There is added information in doing also an SNP test with EA. First, sometimes the prediction is incorrect, above all for non-European haplogroups (since the databases are small). Second, the EA SNP test will usually go a little deeper. Eg., the genographic project may tell you E, but EA may tell you a more precise Esomething, etc.

cacio

[vineviz]

Efgen: by the way, I've heard this repeated many times that FTDNA tests for specific locations outside HVR to determine haplogroups, but I have seen too many recent mis-assignments for this to be true. May be they do this for European haplogroups like H or T, but not in general.

FTDNA runs a multiplex haplogroup panel in addition to their HVR panels on every sample. This multiplex test covers about 20 haplogroup, which is many but certainly not all. If someone's mtDNA haplogroup is not unambiguously determined by the multiplex, then FTDNA must still rely on the HVR1 and/or HVR2 results for a prediction.

In short, FTDNA promises to run the mtDNA multiplex test but cannot guarantee that the test alone will produce an unambiguous haplogroup assignment for everyone.

http://www.familytreedna.com/hclade2.html