Interpreting autosomal results correctly?

**tomcat** · 8 July 2007, 08:01 AM

Originally posted by litlnemo

....
So, these are the conclusions I've drawn so far:

The results suggest my principal ancestries are likely to be Norwegian/Swedish, Scottish, and other British Isles including Ireland. (My mom's side is Norwegian/Swedish with some English, and my dad's side is Scottish/English, so this matches exactly what my researched genealogy would imply.)

However, there are a lot of Iberian and North African results. The Iberian could reflect "deep ancestry," but how should I interpret Omnipop's North African matches? And the Native American (eastern US) matches that show up on the map? (My mom's family always said we had Native American ancestry, but so far I've seen no sign of it in the genealogy.) I would tend to think that is from European admixture -- am I right?

Comments are welcome.

(My mtDNA, incidentally, is J*, from Norway.)

Congratulations on getting results that conform, for the most part, to known genealogy!

You could simply ignore the Iberian and North African and Native American matches as they all fall below the top matches to known ancestries. 13-marker CODIS matches are like 12-marker Y-DNA matches, one can have a lot of them and a lot of those will prove irrelevant.

CODIS profile-matching was not developed for the purposes to which we and Tribes are putting it, the system was developed for forensics to reliably identify individuals from one another. Having said that, it is remarkable that, in your case, results can be spot-on and, as in my case, they are more subtly appropriate.

**litlnemo** · 9 July 2007, 08:13 PM

Originally posted by tomcat

You could simply ignore the Iberian and North African and Native American matches as they all fall below the top matches to known ancestries. 13-marker CODIS matches are like 12-marker Y-DNA matches, one can have a lot of them and a lot of those will prove irrelevant.

I only used the "green" markers because otherwise I lost so many populations it didn't seem to be useful. For example, I lost the Scottish population entirely when adding more markers. So this list is based on 9. Once you start adding more markers, the Lumbee population moves up the list, but I think this is just because the others aren't part of the equation any more... yes?

**tomcat** · 10 July 2007, 08:47 AM

Originally posted by litlnemo

I only used the "green" markers because otherwise I lost so many populations it didn't seem to be useful. For example, I lost the Scottish population entirely when adding more markers. So this list is based on 9. Once you start adding more markers, the Lumbee population moves up the list, but I think this is just because the others aren't part of the equation any more... yes?

You can see population datasets under the Data tab at the foot of the Omnipop page. It could be that the Scottish data is based on 9 markers, the original CODIS standard, and cannot figure into calculations based on 13 markers. If you just use 9 markers you have access to all populations in Omnipop but only on the basis of the 9 original markers.

**tomcat** · 10 July 2007, 10:44 AM

Originally posted by litlnemo

... there are a lot of Iberian and North African results. The Iberian could reflect "deep ancestry," but how should I interpret Omnipop's North African matches? And the Native American (eastern US) matches that show up on the map? (My mom's family always said we had Native American ancestry, but so far I've seen no sign of it in the genealogy.) I would tend to think that is from European admixture -- am I right? ...

I haven't an answer to this specific question but can propose a test. Go to the Data in Omnipop and extract a purely Native American partial profile (a single predominant allele at each locus) and then run that partial in Omnipop to see what comes up. If you get a lot of North African matches along with Native American then you know that a partial Native American profile resembles those of North Africans. Of course, you could do it the other way and start with an extracted North African profile. And you could do the same with the Scottish, or any dataset, in Omnipop.

**jdanel** · 10 July 2007, 03:31 PM

Big E means no info?

I ran omnipop on my results. The lowest E number comes back 17 and it is only one. All the others are E18 to E26.

The frequency diagram is a picket fence with just one picket on the far right end.

These seem to me to be extremely low numbers. It suggests that my results do not resemble ANY similar group (yet).

And it suggests that the listed results are really nothing more than "background noise".

Does this sound right to those of you who are the experts?

**Johnserrat** · 10 July 2007, 03:52 PM

Originally posted by jdanel

I ran omnipop on my results. The lowest E number comes back 17 and it is only one. All the others are E18 to E26.

The frequency diagram is a picket fence with just one picket on the far right end.

These seem to me to be extremely low numbers. It suggests that my results do not resemble ANY similar group (yet).

And it suggests that the listed results are really nothing more than "background noise".

Does this sound right to those of you who are the experts?

Try just entering the CODIS numbers. Many of the population samples do not go beyond the CODIS markers and, as a result, will be excluded when you add additional markers. This should give you more results.

John

**jdanel** · 10 July 2007, 04:03 PM

green dots on omnipop map

Does onmipop color the dots based solely on rank order?

Like maybe the top 20% of the matches get bright green, the next get dark green, the middle group gets brown and the lower ones get reds.

This color scheme seems independent of the actual strength of the info, which in my case, seems nil.

With my markers mostly E24 and higher, should not all my dots be red?

**jdanel** · 10 July 2007, 04:06 PM

To John

I did it both ways, 9 and 13. Same result.

**jdanel** · 10 July 2007, 04:41 PM

Nul result on ENSFI

And on ENSFI, all my Es are 11 or 12. Again this looks like a nul result.

My wife is enjoying this immensely. She is humming the "Twilight Zone" theme and muttering about Roswell and the aliens. Comments like "I told you so."

She couldn't be right, could she? Alien DNA and all that? (yes, I'm joking, but she may not be.)

**tomcat** · 10 July 2007, 08:42 PM

Originally posted by jdanel

I ran omnipop on my results. The lowest E number comes back 17 and it is only one. All the others are E18 to E26.

The frequency diagram is a picket fence with just one picket on the far right end.

These seem to me to be extremely low numbers. It suggests that my results do not resemble ANY similar group (yet).

And it suggests that the listed results are really nothing more than "background noise".

Does this sound right to those of you who are the experts?

In this game ALL the numbers are .000000BIG or all the probabilities are .0000000small. Nevertheless some are more probable than others. If you know the group to which you ought to match and that group is not part of the Omnipop or ENFSI db's then you cannot expect a match. And if you are admixed (like many of us) you cannot expect a stellar match. And if you are extra-terrestrial (like many of us wish we were), CALL HOME, get an e-addy for the ET DNA calculator and please post it here!

**Maria_W** · 10 July 2007, 08:47 PM

ET, phone home!

Tom.
Thats so cute!

Maria

**Maria_W** · 10 July 2007, 09:33 PM

Alien DNA?

There are alot of people that swear that they have alien DNA. This is no joke. They believe that they are products of matings between humans and aliens. They call themselves Star People. One couple who promoted this theory was Brad and Francine Steiger. They talk about being visited by Aliens. So....
Maria

**tomcat** · 10 July 2007, 10:06 PM

Googled 'star people dna' =

http://www.google.com/search?hl=en&q=star+people+dna&btnG=Google+Search

Go for it. Please report back. It can be soooo lonely on FTDNA Forums

**rainbow** · 11 July 2007, 06:03 AM

Originally posted by tomcat

Googled 'star people dna' =

http://www.google.com/search?hl=en&q=star+people+dna&btnG=Google+Search

Go for it. Please report back. It can be soooo lonely on FTDNA Forums

I never heard of Star People before. I'm now looking at the first one from 'google'

That may explain why I match Mozambique, Uyghur 2 (Xinjiang, China), Alaskan Athabaskan, Piemonte, Italy, West Polynesian, Kathmandu, Nepal, & Maori, etc....