DNA Tribes Upgrade- Big Difference

[Brunetmj]

I was cutting and pasting so this time I will type directly.
I just wanted to comment on DKF post. It sort of implies theses tests are worthless unless conducted at a much higher level. If you are addressing the fact that some may be reading to much into these tests without sufficient empirical evidence then I would tend to agree. However information can be used in very different ways for different purposes. If for example traditional genealogy is stuck than these tests can point in a direction of where to look. By choosing a few different tests at an affordable price say DNA tribes and a YDNA test would probably give some evidence for where to look. Just a thought.

[Javelin]

I still do not know why anyone would think it is better to order 20-odd STRs that come from various of their autosomes, and therefore various different ancestors possibly separated by whole continents, for more than the cost of 1 million SNPs at 23andme. How are you supposed to know "where to look" with DNA Tribes if you can get such conflicting information as being Sudanese, Navajo, and Welsh in your top dozen matches?

[Brunetmj]

I still do not know why anyone would think it is better to order 20-odd STRs that come from various of their autosomes

I know in my case it was a lack of knowledge.
Having said that it did provide some very unexpected results which led to further investigation. An expensive lesson but I have made worse mistakes

[rainbow]

I am happy with my DNA Tribes 27 STR marker results. It is very informative. The FBI and other agencies use STR markers.


DKF's Saudi Arabia and Yemen results are probably from the 9 marker test from 2006. Or fewer than 9?

[rainbow]

Part A has the Autosomal STR Profile. The 27 markers have 2 alleles each. I can also search for more info on each of the 54 alleles.

http://www.cstl.nist.gov.biotech/strbase/

http://dnatribes.com/sampleresults.html

[MMaddi]

I am happy with my DNA Tribes 27 STR marker results. It is very informative. The FBI and other agencies use STR markers.

I keep seeing this (FBI use of DNA Tribes-type technology) repeated by those who advocate DNA Tribes and its test results. It certainly is true that the FBI and law enforcement crime labs use CODIS markers in their work; just watch one of the "CSI" shows and that's clear.

However, are you aware that law enforcement agencies use CODIS markers to catch a specific individual who left DNA at a crime scene, not to find out the ethnic background of the suspect? The police want to match up lab results with a specific individual who matches the crime scene DNA. If they were to "get such conflicting information as being Sudanese, Navajo, and Welsh in your top dozen matches," to quote Javelin above, from a DNA Tribes-type test, they would be sent on a wild goose chase and also subject themselves to civil lawsuits for racial/ethnic profiling!

[rainbow]

I feel I've being insulted and bullied over the years on the two forums (here and dna-forums.org)by a psychology teacher Faux (self-described 'scientist') and his cronies.
I am not the one here who is/was connected to/investor of a dna company. Did it used to be deCODEme and now it's 23andme?


Here is a link to the FAQ on DNA Tribes:

http://www.dnatribes.com/faq.html

It includes the following:
DNA Tribes' propietary analysis was developed with Dr. Eduardas Valaitis (Ph.D. in Statistics, Yale University 2005), whose expertise in multivariate classification allows us to provide the most detailed and comprehensive regional analysis anywhere.

[rainbow]

27 marker update. If there is anyone here reading this who has done the 27 marker update, I would hope you aren't intimidated by the bullies, but you probably are.
This is supposed to be an open forum and this thread is for sharing new results.


DFK (and Javelin etc),
Quit criticizing my results. If you want any credibility with me then update your own DNA Tribes to the 2011 27 marker update and critique your own results if you want to prove it's wrong for you. Otherwise, you don't have a leg to stand on.

[Yaffa]

I keep seeing this (FBI use of DNA Tribes-type technology) repeated by those who advocate DNA Tribes and its test results. It certainly is true that the FBI and law enforcement crime labs use CODIS markers in their work; just watch one of the "CSI" shows and that's clear.

However, are you aware that law enforcement agencies use CODIS markers to catch a specific individual who left DNA at a crime scene, not to find out the ethnic background of the suspect? The police want to match up lab results with a specific individual who matches the crime scene DNA. If they were to "get such conflicting information as being Sudanese, Navajo, and Welsh in your top dozen matches," to quote Javelin above, from a DNA Tribes-type test, they would be sent on a wild goose chase and also subject themselves to civil lawsuits for racial/ethnic profiling!

I was told the same thing and have never taken this test. Isn't this what they call a fingerprint test ? DNA markers they use works like a fingerprint left at the seen of a crime which have nothing to do with ethnic back round?

[Javelin]

rainbow, credibility is based on scientific rigor and methodology, not on having paid money for a service. If you spend $20 at a midway ring toss, you don't become an expert on glass bottles, you're just $20 poorer. But if you can't approach gentic testing with the rational detachment that science requires, I have nothing further to say.

This is supposed to be an open forum and this thread is for sharing new results.


DFK (and Javelin etc),
Quit criticizing my results. If you want any credibility with me then update your own DNA Tribes to the 2011 27 marker update and critique your own results if you want to prove it's wrong for you. Otherwise, you don't have a leg to stand on.

[DKF]

Ok, lets go at this from another angle - two actually.

1) With all of the immense data relating to world - wide groups with high specificity (e.g., down to the tribal level such as Athabaskans), why is it that there is not a single population geneticist on the planet who is using this 27 marker information to publish in a peer reviewed journal. I can answer that. Because there is little to no ancestral informational content with which to work. For example 16 repeats at marker Z is widely distributed around the world. The best that can be hoped for is a rough and inexact assignment to a Continental group - if the person is not admixed.

2) There are 22 autosomes and 27 probes (STR markers). This means that there will be about one data point sampled per chromosome. In my case, if they happen to sample the q end of chromosome 16 they will find no African. If by chance it is the tip of the p end of this chromosome where the marker is positioned they will find one African allele and one European allele (as determined by genome wide scan testing (thousands of markers on each chromosome). Say that the "African" repeat value is 16. However 16 is almost certainly found in other populations in the world, simply at a lesser percentage in the groups. So maybe they will find the small amount of African, but it would be by sheer luck because the African segment ends at 10 Mb from the p tip.

Another marker is say found on chromosome 1 at the q end where both of my parents gave me some sort of European value. Lets say it is 13,17. I know that my father is 100% East Anglian from England, and my mother has about one quarter East Anglian heritage (testing cousins living in England confirms there were no known non parental events in the last 200 years. It so happens that this test does not have any sample from East Anglia so the algorithm will be forced to find something from the available pool. The closest approximation may be the north coast of Spain - not because of any shared heritage, but because of chance.

Hence the output I receive as my "results" will be an untrustworthy array of meaningless locations. I already know from genealogy and genome wide scans that I have only a small amount of African and Native American, and the rest is British and German. Hence, British and German should be at the top of my list (they are not). The minority ancestry is much too small to be picked up by a measure using 27 markers so if I saw for example Haida (West Coast Native Canadian) it is only a romantic fantasy that would lead me to do anything but ignore it. The likelyhood of a 27 marker test picking up anything relating to someone with 1/256 NA ancestry would be akin to winning the power ball lottery.

The reality is that tests such as these have been left in the dust years ago - once chip technology and genome wide scans entered the picture. We will learn a great deal here via using the raw data with methods such as multidimensional scaling to view specific segments of say 15 Mb in length that were inherited from a Native American ancestor. This will be a good introduction to what we can garner from full genome sequencing which is the final frontier (unless one wants to discuss epigenetics and how the experiences of one's ancestors may have "written upon" the expression of the DNA).

Using the raw data from 23andMe testing, Dr. McDonald was able to place my genome on the western border between England and Scotland. That is amazing since about 70% of my genome is from Eastern England, much from Yorkshire. The "pull" from the German and the Irish will mean that only such testing can tell where my genome will roost. If my father had been tested, I am fairly sure Dr. McDonald would have been able to place him somewhere between the Scottish border and Essex. No 27 marker test can remotely offer this specificity.

[Brunetmj]

credibility is based on scientific rigor and methodology, not on having paid money for a service

That's pretty funny just ask Dr Andrew Wakefield. But that's another matter.
What are your opinions on using the very top results from Dnatribes test and then comparing them with y DNA and a mDNA test from this site?

[Javelin]

Thanks! I still wouldn't suggest ordering it, since for about the same price you will get much more insight into your recent ancestry from Family Finder or from 23andme, or from testing one or a few cousins' mtDNA or yDNA that are ancestral to you but on your non-direct lines (e.g. a maternal uncle's Y, or a cousin who shares your father's father's mtDNA line).

But if you have already ordered DNA Tribes and cannot cancel, when the results come in you could at least try to identify which population some of your alleles may have come from with an offsite tool such as EHSTRAFD or the OmniPop spreadsheet.