The Correct Way to Enter OmniPop Data for Autosomal Results

[Arch Yeomans]

From my understanding the correct way to enter your autosomal DNA into OmniPop in order to get more accurate results is to add the first 5 CODIS markers, then map it, erase the first 5 CODIS markers and add the rest of the CODIS markers in, and remap it. Make a comparison of the clusters from both maps and this should give a better idea where your ancestral DNA is most likely to be found. The tighter the clusters between both maps on average would be the best bet (obviously). For frequency results, the lower numbers and lower E number are supposed to be 'closer' to the target than the higher numbers. Somebody please correct me if this is wrong.

[tomcat]

I haven't heard of this heretofor.

What is the rationale behind the procedure of running partial profiles and comparing them?

Do your run 4 and then 5 to get the 9 Green Cell data and then the 4 (CSF, D16, TH01, TPOX) that are supported by more limited data sets?

Do you rearrange the marker combinations? By my reckoning, the 9 Green Cell markers run as triplets would offer 27 combinations - 'manageable.'

[camoe]

Oh...pray tell, where did this methodology come from? I'm anxious to get more pertinent data from my autosomal results.

Cheers

[tomcat]

...
Do you rearrange the marker combinations? By my reckoning, the 9 Green Cell markers run as triplets would offer 27 combinations - 'manageable.'

Correction: 30

[Maria_W]

How?

If this is true then alot of us have been doing it wrong. And why this way? I am already confused just reading the directions.
Maria

[tomcat]

Oh...pray tell, where did this methodology come from? I'm anxious to get more pertinent data from my autosomal results.
Cheers

Me too! I can't speak for Arch Yeomans, but I compose my methodologies and just run them on Omnipop to see what I get. Here's a recent experiment posted on Native American Tribes thread -

"With profiles for myself and three of my sisters in-hand I have been able to identify parental alleles for D18, vWA and FGA and decided to Omnipop partial profiles to see if I could discern to which parent which allele pair belonged.

With just three markers the combinations are few - eight, or four pairs as each combo has a reciprocal. I ignored all admixed/diaspora matches and all out-of-area matches (China etc.). The (top twelve) results were very odd as each pair produced a high-scoring Jewish match and it's mate one or more high-scoring Native matches:

1) Jewish 8.18/03
2) Inupiat 2.32/03, Yupik 4.18/03, Lumbee 8.12/03

3) Jewish 1.46/04
4) Inupiat 1.45/03, Yupik 3.72/03

5) Inupiat 7.74/03, Salish 8.86/03, Salish 1.02/04, Yupik 1.31/04
6) Jewish 3.41/03

7) Yupik 1.47/04
8) Jewish 3.41/03 plus Yupik at 6.02/03

In the hope of resolving the matter, I then added to each profile single alleles for D3 and D5 where sibling homozygotic results indicated that both parents had a common allele:

1) Jewish 2.53/E04 plus PBSO Native Amer 3.20, ABI-ID Native Amer 3.23
2) Inupiat 4.35/E03, Yupik 1.25/04, Lumbee 1.33

3) Jewish 4.52/E04
4) Inupiat 2.72/E03

5) Inupiat 1.45/E04, Salish 2.22 & 2.74, PBSO Native Amer 3.14, Yupik 3.91
6) Inupiat 1.81/E04 and Jewish 1.88

7) Inupiat 2.32/E04, Yupik 4.40
8) Jewish 1.05/E04 plus Inupiat 1.13

Then I tossed in a TPOX allele common to both parents. As this closed me out of the Jewish sample in Omnipop that does not use TPOX data the (Jewish) in the following is just place-holding previous results:

1) (Jewish) PBSO Native American 3.67/E04, MN Native American 4.05
2) Inupiat 8.78/E03 Yupik 2.98/E04

3) (Jewish)
4) Inupiat 5.49/E03, Lumbee 6.13, MN Native American 2.40/E04

5) Salish 2.22/E04 & 2.71, Inupiat 2.93, PBSO Native 3.60, MN Native 5.84
6) (Jewish)

7) Inupiat 4.69/E04, Lumbee 6.58, Salish 6.76 & 6.79, PBSO Native 6.96
8) (Jewish)

Comments and suggestions invited. I want to make a most likely attribution of these profiles to my parents - 1-2 or 3-4 or 5-6 or 7-8."

[tomcat]

Notes to the above:

These Jewish and Native matches are not always the top-scoring matches (although Natives take first place in 6 out of 8 runs) but the Jewish and Native matches are, for me, the most relevant of the top 12 matches (barring admixed/diaspora and genealogically-irrelevant out-of-area matches).

Anyhow, I thought of it and did it. It was a failure, I guess, but an exceedingly peculiar failure. I posted it to see if anyone had an idea of how to resolve it. I wanted Omnipop to give me a clear result but instead Omnipop gave me a choice. I don't want to choose.

[Arch Yeomans]

No rationale, just what I found on the web...

I haven't heard of this heretofor.

What is the rationale behind the procedure of running partial profiles and comparing them?

Do your run 4 and then 5 to get the 9 Green Cell data and then the 4 (CSF, D16, TH01, TPOX) that are supported by more limited data sets?

Do you rearrange the marker combinations? By my reckoning, the 9 Green Cell markers run as triplets would offer 27 combinations - 'manageable.'

I can't remember which board I found this information on. I believe it had something to do with African American ancestry. It seemed quite legitimate and referenced to Dr. Thomas Khran (?) and his advice on how to get some sort of an idea of where your immediate, or fairly immediate ancestry came from. I tried it and it didn't make much sense at first, but when I compared only the CODIS markers and followed the directions; it seemed to cluster quite highly around Iberia (confirming my haplogroup R1b1c6?) and as well cluster highly around Great Britain (confirming a fairly recent migration to North America within the last couple of hundred of years, e.g., 1800s). Dammit I'm not a doctor Jim! But try it and see what results you get.

I also entered all the autosomal data and a very high percentage showed Hispanic, Caucasian, and Madeira Archipelago; as well a lot of Indian (which I found a bit weird). And some Native American and African American (not much). I posted my results at DNA-Forums.Org under the Autosomal Directory. Sorry I can't answer the question about arrangements as I'm still learning this OminPop program to see whether it's as good as it's advertised.

[Arch Yeomans]

Instructions

If this is true then alot of us have been doing it wrong. And why this way? I am already confused just reading the directions.
Maria

Enter your allele values for the first 5 CODIS markers (fields highlighted green) these are the ones I believe the FBI use. Then map it. Look at how everything is clustered.

Now, erase those values and enter the allele values in the remaining CODIS markers and remap. Look how everything is clustered.

Now print out your result sheets from the mappings and overlay one sheet over the other. This is allegedly a higher probability where your recent ancestry is from.

Mine seem to be accurate, but not as accurate as I would like of course. I've also entered all the data from panel 1 in OmniPop and it's still close to the mark. It shows Caucasian, Hispanic, Madeira Archipelago, Tunisia, Albania (Northern Italy). Where the Indian stuff comes from I don't know, but I do know my maternal great great Grandmother was Native American and the Native American results in OminPop may identify her tribe.

ENFSI is pretty good too.