PDA

View Full Version : Is 23andme the most accurate DNA testing company?



Max
12-12-11, 02:33
Is 23andme the most accurate DNA testing company?

sparkey
12-12-11, 19:40
It's accurate, but so are the other companies. People who test with 23andMe usually have a different goal than people who test with Family Tree DNA. Why do you ask?

Max
12-12-11, 19:49
It's accurate, but so are the other companies. People who test with 23andMe usually have a different goal than people who test with Family Tree DNA. Why do you ask?
I want to get tested but I couldn't decide which company.:) People keep recommending 23andMe to me, but what's the difference with ftdna?

sparkey
12-12-11, 20:05
I want to get tested but I couldn't decide which company.:) People keep recommending 23andMe to me, but what's the difference with ftdna?

OK, which of these would be something you'd say?:

(1) I'm stuck on my direct patriline on my family tree, and want to see who I match with along it. I'm also interested in being very specific about my deep ancestry along my patriline (and maybe also my matriline) and want to contribute to Y-DNA projects. I want to easily do additional tests down the line if I'm unsatisfied with my results.

(2) I want a broad, affordable genetic test that will tell me all the basics, including a bit about my patriline, a bit about my matriline, and a good deal about everything in-between. I want to be able to contribute to autosomal DNA projects and find out who I match with in general, what countries and ethnic groups are closest to me, etc.

If (1) go with FTDNA. If (2) go with 23andMe. (1) fit me better but most here were in situation (2) so 23andMe tends to be more popular here.

Max
13-12-11, 00:02
OK, which of these would be something you'd say?:

(1) I'm stuck on my direct patriline on my family tree, and want to see who I match with along it. I'm also interested in being very specific about my deep ancestry along my patriline (and maybe also my matriline) and want to contribute to Y-DNA projects. I want to easily do additional tests down the line if I'm unsatisfied with my results.

(2) I want a broad, affordable genetic test that will tell me all the basics, including a bit about my patriline, a bit about my matriline, and a good deal about everything in-between. I want to be able to contribute to autosomal DNA projects and find out who I match with in general, what countries and ethnic groups are closest to me, etc.

If (1) go with FTDNA. If (2) go with 23andMe. (1) fit me better but most here were in situation (2) so 23andMe tends to be more popular here.
:) Thanks for the info, number 2 sounds better for me :good_job:

Dorianfinder
13-12-11, 00:11
OK, which of these would be something you'd say?:

(1) I'm stuck on my direct patriline on my family tree, and want to see who I match with along it. I'm also interested in being very specific about my deep ancestry along my patriline (and maybe also my matriline) and want to contribute to Y-DNA projects. I want to easily do additional tests down the line if I'm unsatisfied with my results.

(2) I want a broad, affordable genetic test that will tell me all the basics, including a bit about my patriline, a bit about my matriline, and a good deal about everything in-between. I want to be able to contribute to autosomal DNA projects and find out who I match with in general, what countries and ethnic groups are closest to me, etc.

If (1) go with FTDNA. If (2) go with 23andMe. (1) fit me better but most here were in situation (2) so 23andMe tends to be more popular here.

For the record sparkey, FTDNA's Family Finder also provides the opportunity to contribute to autosomal DNA projects and it is cheaper as there are no subscription costs. Also, FTDNA provides individual STR-marker values and allows for predicted haplogroup results with the cheapest Y-STR marker tests. The only advantage that 23andMe has over FTDNA are their medical predictions, however this is a controversial area of commercial genetic testing.

sparkey
13-12-11, 00:25
For the record sparkey, FTDNA's Family Finder also provides the opportunity to contribute to autosomal DNA projects and it is cheaper as there are no subscription costs. Also, FTDNA provides individual STR-marker values and allows for predicted haplogroup results with the cheapest Y-STR marker tests. The only advantage that 23andMe has over FTDNA at the current moment is in the medical predictions.

That's mostly true, although, obviously, if you go for FTDNA's Family Finder, you will still need to get their Y-DNA and mtDNA tests entirely separately. Theoretically you can get more from FTDNA, but it will rack up a cost. Are you sure about FTDNA being cheaper? FTDNA's FamilyFinder+YDNA12 is $288 right now, compared with 23andMe's basic package at $207.

I'm a fan of Y-DNA projects myself, so for me, it's FTDNA, no contest, anyway.

Maciamo
13-12-11, 13:56
Keep in mind that for a similar price 23andMe tests 1 million SNP's, while FTDNA's Family Finder only tests half (563,800 SNPs). The Family Finder doesn't test the Y-chromosome at all, so additional tests are still required (and FTDNA's custom SNP tests for deep subclades aren't cheap, to say the least).

I would recommend 23andMe at first. After that, if you feel the need after that to join a Y-DNA Project you can still order an STR test, but I feel that they are quite expensive for the information they provide. Some Y-haplogroups benefit more from knowing STR values than others. This is especially the case of haplogroups that still have a lot of unknown SNP's and have active researchers for the STR. I think that so far it can be said that haplogroup I is the European haplogroup for which STR"s provide the most information in addition to SNP's.

Dorianfinder
13-12-11, 16:00
Keep in mind that for a similar price 23andMe tests 1 million SNP's, while FTDNA's Family Finder only tests half (563,800 SNPs). The Family Finder doesn't test the Y-chromosome at all, so additional tests are still required (and FTDNA's custom SNP tests for deep subclades aren't cheap, to say the least).

I would recommend 23andMe at first. After that, if you feel the need after that to join a Y-DNA Project you can still order an STR test, but I feel that they are quite expensive for the information they provide. Some Y-haplogroups benefit more from knowing STR values than others. This is especially the case of haplogroups that still have a lot of unknown SNP's and have active researchers for the STR. I think that so far it can be said that haplogroup I is the European haplogroup for which STR"s provide the most information in addition to SNP's.

Maciamo, your info regarding FTDNA has been outdated for quite some time now. FTDNA's Family Finder tests 710 000 ancestral SNPs using the v2 Illumina chip. However, the Dodecad project has found no additional advantage in testing more than 150 000 autosomal (ancestral) SNPs. FTDNA's X-chromosome analysis tests more SNPs than 23andMe's RR X-chromosome analysis.

For genealogists, STR-values are important and 23andMe does not provide the same quality of service as currently provided by FTDNA. The most significant advantage of FTDNA's service is that there is no subscription required and customer service is exceptional. In addition, FTDNA keeps your DNA for 25yrs in which time your results are compared free of charge, and you receive feedback as new matches are made. FTDNA allows one to choose exactly what one needs whereas 23andMe provides extras that are not asked for by clients. FTDNA is a better service, customers pay once for a product, and STR-values are made available to clients.

Lastly, 23andMe does not provide tech support and make provision for Family Surname Groups or Geographic Projects, with numerous features and STR info. FTDNA has introduced many genealogists to the field and provided them with support to grow their understanding and contribute to ancestral research as amateur genetic genealogists. 23andMe lacks the customer-interoperability available to FTDNA clients!

Personally, I would like to see more from 23andMe's user-interface, all in the interests of competition and R&D. The health predictions 23andMe is known for have attracted unnecessary bad press/controversy to the field of genetic genealogy. IMHO it has set the image and growth back quite some. I don't like the confounding aspects and continuous bad press surrounding confidentiality, this is a non-issue regarding FTDNA's autosomal analysis as NO health-informative SNPs have been included (jolly good thing). Ancestry and health are separate fields, 23andMe has managed to muddle the two rather successfully. Essentially, 23andMe ancestral matches with other customers are over-regulated for fears of health-related information being made public without the necessary consent. Poor planning by 23andMe has forced the company into a protocol that is unfriendly to genetic genealogists at the very core of the ancestry product's service, genealogical research.

Maciamo
13-12-11, 19:33
Maciamo, your info regarding FTDNA has been outdated for quite some time now. FTDNA's Family Finder tests 710 000 ancestral SNPs using the v2 Illumina chip. However, the Dodecad project has found no additional advantage in testing more than 150 000 autosomal (ancestral) SNPs. FTDNA's X-chromosome analysis tests more SNPs than 23andMe's RR X-chromosome analysis.

Yes, you are right that it's 710,000 now, not 563,800 SNPs. But that's not a big improvement.

It depends again what you are interested in. I have zero interest in finding distant cousins. I already know all my first and second cousins, and have a family tree with exactly all my third and fourth cousins, which I don't care much about anyway.

On the other hand I care a lot about medical genetics, and the more SNP's the better.


For genealogists, STR-values are important and 23andMe does not provide the same quality of service as currently provided by FTDNA.

But that's a completely different service from SNP tests like the Family Finder or 23andMe. Nothing prevents anybody from taking both tests, as they are completely separate anyway. But when it comes to SNP's 23andMe is superior.

FTDNA claim that comparing STR can work to complete/confirm your family tree, but you still need to get relatives to test. I have contacted dozens of people with the same surname as mine and never got anybody to take the test, even when I proposed to pay for it. These were generally genealogists, but they don't trust DNA tests, especially if they are done by commercial companies in the USA. I am still lucky to have a reasonably common surname, since the average in Belgium is 30 persons per surname (http://www.eupedia.com/europe/european_family_names.shtml). In may branches of my family tree there are family names that are extinct (even going back a 4 or 5 generations) or that have very very few members alive today (e.g. 10 or 20 in the world). So I have never found any potential in testing even a single line of my huge family tree (going back to the 13th century) simply because there aren't enough people interested in genetic genealogy in my country. It is the case in many countries in continental Europe. Only English (and perhaps German) speaking countries have enough participants to find STR tests useful for genealogy.



The most significant advantage of FTDNA's service is that there is no subscription required and customer service is exceptional.

You can buy 23andMe without subscription. From my personal experience FTDNA's customer service sucks. I have actually ordered a few tests from them (including extension of STR panels and custom SNP tests before 23andMe existed). Once it took 6 months to get the results, but that was nothing compared to the time when it took over 1 year ! I contacted Bennett Greenspan directly, but it didn't speed up things much. They just found some lame excuses.



In addition, FTDNA keeps your DNA for 25yrs in which time your results are compared free of charge, and you receive feedback as new matches are made.

I know. I still get regular emails claiming that they found "matches" when in fact none of those people could possibly be related to me in genealogical times (let's say within 500 years).


FTDNA allows one to choose exactly what one needs whereas 23andMe provides extras that are not asked for by clients

You are kidding right ? How do you choose your SNP's in the Family Finder ? It's exactly the same as 23andMe. You get a full package of hundreds of thousands of SNP's. It's true that FTDNA also propose to test single custom SNP's for Y-DNA. But ordering single SNP's is extremely expensive and therefore in no one's best interest. There is absolutely no good reason not to want more SNP's for the same price.



Lastly, 23andMe does not provide tech support

What kind of tech support ?


The health predictions 23andMe is known for have attracted unnecessary bad press/controversy to the field of genetic genealogy.

Why is that ? That's far more scientific than genetic genealogy. I truly think that FTDNA is scamming people when they say that they can find matrilineal relatives by testing mtDNA. I have hundreds of matches on my full mtDNA sequence and none are related within at least 15 generations, whereas FTDNA claims that they could be related within a few generations.


IMHO it has set the image and growth back quite some. I don't like the confounding aspects and continuous bad press surrounding confidentiality, this is a non-issue regarding FTDNA's autosomal analysis as NO health-informative SNPs have been included (jolly good thing). Ancestry and health are separate fields, 23andMe has managed to muddle the two rather successfully. Essentially, 23andMe ancestral matches with other customers are over-regulated for fears of health-related information being made public without the necessary consent.

Confidentially is more of an issue with FTDNA if one joins projects or post their data on Y-search or mitosearch, as surnames are normally used (a minority hides it, justly for confidentiality reason as anybody can see it). Besides it's not because FTDNA's Family Finder doesn't report health risks that it does not contain the information. By sending your data from FTDNA's Family Finder to autosomal projects you are sending just as much information about your health risks as with the 23andMe data.

Dorianfinder
14-12-11, 00:50
But that's a completely different service from SNP tests like the Family Finder or 23andMe. Nothing prevents anybody from taking both tests, as they are completely separate anyway. But when it comes to SNP's 23andMe is superior.

FTDNA provides the WTY, an advanced deep clade test, to uncover new SNPs. 23andMe's equivalent is similar to the deep clade test provided by FTDNA. FTDNA consciously decided to not include medical-informative SNPs as customers often send their data files for analysis to other companies or individuals. This is why FTDNA does not have confidentiality issues, the 'problem' SNPs have been removed to safeguard their customers. The v3 chip used by 23andMe has numerous no-calls and no additional ancestral information is found, superior in terms of quantity, not quality?


FTDNA claim that comparing STR can work to complete/confirm your family tree, but you still need to get relatives to test. ... Only English (and perhaps German) speaking countries have enough participants to find STR tests useful for genealogy.

STR-marker values are very significant when placing individuals within clusters that are often as useful as SNPs if not more in certain cases. Surnames are a poor indication of ancestry in general. FTDNA provides over 100 STR-marker values that, together with FTDNA's yDNA deep clade test, provide superior analysis when comparing males with different last names that share a similar haplogroup/subclade. TMRCA is based on STR information and FTDNA gives customers the opportunity to test more STRs, giving more accurate TMRCA estimates.


I know. I still get regular emails claiming that they found "matches" when in fact none of those people could possibly be related to me in genealogical times (let's say within 500 years).

FTDNA only provides matches at the setting which you choose. If you got 12-marker matches that were largely insignificant then why did you not change the setting to 37-markers or 67-marker matches. Matches at these levels are significant.


You are kidding right ? How do you choose your SNP's in the Family Finder ? It's exactly the same as 23andMe. You get a full package of hundreds of thousands of SNP's. It's true that FTDNA also propose to test single custom SNP's for Y-DNA. But ordering single SNP's is extremely expensive and therefore in no one's best interest. There is absolutely no good reason not to want more SNP's for the same price.

Family Finder does not include health-related SNPs, you don't choose them as they are not available. 23andMe finds these SNPs whether a customer only wants ancestral SNPs or not. In contrast to 23andMe's mix of health-related and ancestral SNPs in its Relative Finder, FTDNA provides customers with a purely ancestral autosomal test in its Family Finder product, and if a customer would like to find out about health there are individual SNPs that one can order in FTDNA's factoids section.

FTDNA is superior in yDNA analysis as it provides more than 100 STR-marker values, a Deep Clade test, a Walk-Through-The-Y test, and individual SNP tests at $29 a pop. I managed to save by testing for individual SNPs and inferring subclade membership accordingly. 23andMe does not allow for this.


What kind of tech support ?

FTDNA provides customers with the opportunity to upload their gedcom files and allows any customer to create a project using GAP 2.0. The feedback to problems of a technical nature is quick and easily accessible.


Why is that ? That's far more scientific than genetic genealogy. I truly think that FTDNA is scamming people when they say that they can find matrilineal relatives by testing mtDNA. I have hundreds of matches on my full mtDNA sequence and none are related within at least 15 generations, whereas FTDNA claims that they could be related within a few generations.

The Family Finder/full mtdna sequence combination is a powerful tool. FTDNA's gedcom upload capability and the shere number of genealogist customers at FTDNA make a match more useful than 23andMe at this stage. FTDNA does not claim anything as a certainty, this is a similar issue regarding 23andMe's health predictions. FTDNA's mtdna analysis will provide more detailed X-chromosome analysis as part of the Family Finder product. It provides a 'deep clade' mtdna haplogroup test, allowing some haplogroups to test for individual SNPs and therefore go further downstream than 23andMe can.


Confidentially is more of an issue with FTDNA if one joins projects or post their data on Y-search or mitosearch, as surnames are normally used (a minority hides it, justly for confidentiality reason as anybody can see it). Besides it's not because FTDNA's Family Finder doesn't report health risks that it does not contain the information. By sending your data from FTDNA's Family Finder to autosomal projects you are sending just as much information about your health risks as with the 23andMe data.

Confidentiality is a medical-legal issue. No medical info is to be found within FTDNA's y-DNA information provided to customers. Consequently, STR-marker information cannot be used to the detriment of a customer other than make a NPE (non-parental event) public. On the other hand, 23andMe's Relative Finder data files are full of medical-info and include SNPs with very sensitive information. This is a problem as many 23andMe customers have emailed their data files for analyses without legally safeguarding themselves from possible data mining by pharmaceutical and other related companies (for research purposes).

Maciamo
14-12-11, 01:07
Dorianfinder, so basically all you say is that FTDNA is better for genetic genealogy, which is obvious since 23andMe is not designed for genetic genealogy. What's the point in criticising 23andMe for not having a tech support to upload gedcom file or providing TMRCA since it's not their business. It's like saying that you don't like a car because it's not a boat and you really just wanted a boat.

As for the possibility to test new Y-chromosomal SNP's at FTDNA, that may not be tested by 23andMe, it may be interesting in some cases, but I still think that it is too expensive compared to the additional information it provides. For example there has been over 40 new SNP's downstream of R1b-L11 identified this year, but very few provide real information about ancestry except the fact that you belong to this or that new subclade. It may be useful in a few years once we know more about these subclades, but by then they will probably be on the 23andMe chip too.

Regarding the medical SNP's, I absolutely don't understand why you or anybody else would not want to know about their health when given the opportunity. It sounds like an extreme case of the ostrich policy.


No medical info is to be found within FTDNA's y-DNA information provided to customers.

I wrote FTDNA's Family Finder. There is no way they can test 710,000 SNPs without any one of them having medically relevant information. And even if they don't know any medical associations doesn't mean that there isn't any (new studies are published all the time).

By the way, your ultra-defensive and evasive attitude to criticism directed at FTDNA make me think that you work for FTDNA. Do you ?


On the other hand, 23andMe's Relative Finder data files are full of medical-info and include SNPs with very sensitive information. This is a problem as many 23andMe customers have emailed their data files for analyses without legally safeguarding themselves from possible data mining by pharmaceutical and other related companies (for research purposes).

Exactly the same thing about FTDNA's Family Finder data.

Anyway what could pharmaceutical companies do with raw data without knowing the person's health history ? And why would it be a bad thing for pharmaceutical companies to improve their medicines based on genetic data ? It sounds like misguided paranoia. I would rather encourage people to submit (even anonymously) their raw data and medical history to pharmaceutical companies and medical researchers. This is how science progresses.

sparkey
14-12-11, 02:12
I will mention that in my experience, I am very glad I went with FTDNA, and I learned much more with them than I am likely to have learned with 23andMe. Part of that was luck... I happened to have a haplogroup that was good for STR analysis (I2c) and I also happened to have close patrilineal relatives already test through them or through Ancestry.com, which gives results that are more comparable to FTDNA than to 23andMe. The other part was that FTDNA fit my purpose and my focus on my patriline, as I have already described. I actually haven't even done an autosomal test yet... I don't really see a reason to, since I already know so much about my genealogy, and the autosomal patterns that groups within my genealogy tend to fill. If I did, I'm not sure which I'd go with... but then, I've already taken Y-DNA and mtDNA tests.

All considered, I still think I'd tend to recommend 23andMe instead to Max, given his goals. And this technical discussion is likely to be more than Max wanted here.

Dorianfinder
14-12-11, 03:20
Dorianfinder, so basically all you say is that FTDNA is better for genetic genealogy, which is obvious since 23andMe is not designed for genetic genealogy. What's the point in criticising 23andMe for not having a tech support to upload gedcom file or providing TMRCA since it's not their business. It's like saying that you don't like a car because it's not a boat and you really just wanted a boat.

This thread is entitled 'Is 23andMe the most accurate DNA testing company?' hence I refer to accuracy, ancestry and autosomal testing. FTDNA's accuracy regarding yDNA and mtdna testing is better whereas its individual SNP testing, Walk-Through-The-y and overall support to its customers makes for better accuracy in interpreting and reaching further downstream in both y and mtdna terms.

If a customer wants accuracy and is looking at yDNA, FTDNA is the 1st choice. If looking at mtdna, FTDNA is 1st choice. If looking at autosomal testing for ancestral purposes, FTDNA would be my recommendation as there are purely ancestral SNPs and no sensitive information that may potentially cause a customer problems. Regarding accuracy, I have already said that no additional benefits are derived from testing more than 150 000 SNPs.

If you would like to test for medical reasons then I would recommend 23andMe, however the accuracy or correlation between prediction and actual long-term manifestation of predicted conditions are not proven and are considered highly speculative. So if somebody find out they have a 1.2x greater chance of suffering a heart attack, what is the use of this information?


As for the possibility to test new Y-chromosomal SNP's at FTDNA, that may not be tested by 23andMe, it may be interesting in some cases, but I still think that it is too expensive compared to the additional information it provides. For example there has been over 40 new SNP's downstream of R1b-L11 identified this year, but very few provide real information about ancestry except the fact that you belong to this or that new subclade. It may be useful in a few years once we know more about these subclades, but by then they will probably be on the 23andMe chip too.

What would you rather want to know about yourself? That you belong to a specific subclade found in a specific group of individuals or that you have a mildly higher chance of developing a specific form of cancer. Shouldn't you live a healthy life-style regardless and wouldn't most people find health predictions difficult to interpret, process and express to their significant others?


I wrote FTDNA's Family Finder. There is no way they can test 710,000 SNPs without any one of them having medically relevant information. And even if they don't know any medical associations doesn't mean that there isn't any (new studies are published all the time).

You actually referred to y-search and other tools as well. But that's besides the point, FTDNA has removed those autosomal markers used by 23andMe for health-related predictions. This is not the same as 23andMe purposefully testing for these markers and then customers sending their data for ancestral percentages without an option to remove the non-ancestral info from their data files. This is unethical is it not?


By the way, your ultra-defensive and evasive attitude to criticism directed at FTDNA make me think that you work for FTDNA. Do you ?

I am not affiliated with any DNA testing companies. I myself have experienced delays but understand this to be as a result of backlog, if anything I would say this serves the interest of accuracy. Speed is good when one requests support and FTDNA is known for its response efficacy.


Regarding the medical SNP's, I absolutely don't understand why you or anybody else would not want to know about their health when given the opportunity. It sounds like an extreme case of the ostrich policy.


Anyway what could pharmaceutical companies do with raw data without knowing the person's health history ? And why would it be a bad thing for pharmaceutical companies to improve their medicines based on genetic data ? It sounds like misguided paranoia. I would rather encourage people to submit (even anonymously) their raw data and medical history to pharmaceutical companies and medical researchers. This is how science progresses.

What could happen you ask? Should we not ask what is in the best interest of the customer instead? No company or researcher should be privy to this information without following standard ethical protocol! There are numerous medical-legal issues here, least of all consent, confidentiality and access to personal information. You assume that anonymity relates soley to an individual's last name, is this still the case or will this remain the case into the foreseeable future? Am I paranoid for considering the potential harmful effects of this kind of testing? What would you feel like if you found out that you carried the SNP for early-onset dementia or Alzheimers? Would you want to know whether your children carry it and how would your life be effected/affected? If your medical information was used by a company without your knowledge, would you like it?

You have found it difficult getting people tested right?, even when providing the funds yourself, because most people inherently know that there is a risk involved, the risk is personal and often has to do with mental health aspects of finding out about oneself. Like it or not, this is the norm rather than the exception. So if I'm paranoid then so are most people.

Maciamo
14-12-11, 12:16
This thread is entitled 'Is 23andMe the most accurate DNA testing company?' hence I refer to accuracy, ancestry and autosomal testing. FTDNA's accuracy regarding yDNA and mtdna testing is better whereas its individual SNP testing, Walk-Through-The-y and overall support to its customers makes for better accuracy in interpreting and reaching further downstream in both y and mtdna terms.

You are basically dodging my comments. You are not answering nor even considering my point of view. You just push on with the same pro-FTDNA propaganda as if your life depended on it.

I didn't reply to this thread at first when I saw the title question because it seemed like a dumb question to me. Unless one company has more no-calls or mis-reported results, all of them have the same level of accuracy. Note that both 23andMe and FTDNA FF use the same Illumina OmniExpress (http://www.illumina.com/products/human_omni_express.ilmn) platform, and therefore equally accurate. This is the only valid answer to the question of accuracy. FTDNA just customized their chip to remove Y-DNA and some other SNP's so that they could keep charging customers for separate tests.



If you would like to test for medical reasons then I would recommend 23andMe, however the accuracy or correlation between prediction and actual long-term manifestation of predicted conditions are not proven and are considered highly speculative. So if somebody find out they have a 1.2x greater chance of suffering a heart attack, what is the use of this information?

You are confusing accuracy with quantity of information. A calculator that can only do basic additions, subtractions, multiplications and divisions is accurate. A PC that can do all that plus plenty of more things is just as accurate but will be more informative.

FTDNA has five kinds of tests :

1) an SNP test like 23andMe (but less informative than 23andMe because it has 290,000 SNP's less)
2) various STR tests for Y-DNA, useful mostly for genealogists who have relatives who tested
3) custom SNP tests for Y-DNA, which are useful only for people who want to know which brand new deep subclade they belong to.
4) HVR tests for mtDNA which are highly inaccurate and should never be recommended for any purpose. This was one of the first tests I ordered 4 years ago, and the "exact matches" provided by FTDNA turned out not to belong to the same subclades as me at all ! In fact this test couldn't predict my subclade at all, just my mtDNA haplogroup.
5) the full mitochondrial sequence test, which costs more than the 23andMe test and in most cases won't be more informative as 23andMe already tests 90% of the known subclades. Anyway mtDNA is not very informative about ancestry or historical migrations at present, and certainly not useful for genealogy since I have many full sequence matches that are completely unrelated to me, with a long paper trail backing this on both sides.



What would you rather want to know about yourself? That you belong to a specific subclade found in a specific group of individuals or that you have a mildly higher chance of developing a specific form of cancer. Shouldn't you live a healthy life-style regardless and wouldn't most people find health predictions difficult to interpret, process and express to their significant others?


You seem to be unaware that 23andMe also reports carrier status for genetic diseases, for which the likelihood of having the disease or condition is often 100%. This is very useful to know, as some things that could have afflicted you for years can finally be identified (and in some cases treated). Then, there are many other SNPs that give greatly increased/reduced risks of having a disease or condition. A famous one if the APOE4 (http://en.wikipedia.org/wiki/Apolipoprotein_E) version of the APOE gene, which dramatically increases the chances of having Alzheimer's disease.



You actually referred to y-search and other tools as well.

Confidentiality is about knowing your name, not knowing some anonymous medical data. I am less comfortable about sharing my name on Y-search than sharing my 23andMe raw data anonymously with strangers.



What could happen you ask? Should we not ask what is in the best interest of the customer instead? No company or researcher should be privy to this information without following standard ethical protocol!

Are you saying that 23andMe does not follow any ethical protocol ?


You assume that anonymity relates soley to an individual's last name, is this still the case or will this remain the case into the foreseeable future?

What do you mean ? If you are worried about governments collecting DNA from all its citizens, making a huge database, and testing every surface you touched to follow your every movement in your life, that is paranoia (and unrealistic at that, since no government as the financial means to pay for all these tests, nor any interest in doing so. Even if they did I still wouldn't feel threatened as I am a model citizen).


Am I paranoid for considering the potential harmful effects of this kind of testing? What would you feel like if you found out that you carried the SNP for early-onset dementia or Alzheimers? Would you want to know whether your children carry it and how would your life be effected/affected? If your medical information was used by a company without your knowledge, would you like it?

Not only would I want to know all about it, it was also one of my motivations in taking the 23andMe test. Not knowing is what tends to worry me. The first thing I did when I got my results was checking for the conditions/diseases I was most afraid of, and it did appease me.


You have found it difficult getting people tested right?, even when providing the funds yourself, because most people inherently know that there is a risk involved, the risk is personal and often has to do with mental health aspects of finding out about oneself. Like it or not, this is the norm rather than the exception. So if I'm paranoid then so are most people.

And that was just for Y-DNA. But French speakers are often extremely wary of DNA tests. It's a cultural thing, and it explains why there are so few results from France compared to Britain, Germany, Italy or Spain, and why there is such a huge gap in the number of results between French-speaking and Dutch-speaking Belgium.

Knovas
14-12-11, 12:32
So, for what I have read, considering my Paternal haplogroup, ¿Is it possible to obtain more information of it using FTDNA? And ¿what about my MtDNA as well?. Just want to know if it could worth to do so in the future in my case.

Personal curiosity, let's see if somebody versed can help :P

Kardu
14-12-11, 12:43
I've compared FTDNA-s autosomal results to 23andme's using Dodecad K12a calculator and the results are almost identical, 0.01% difference on some components.

Dorianfinder
14-12-11, 15:24
So, for what I have read, considering my Paternal haplogroup, ¿Is it possible to obtain more information of it using FTDNA? And ¿what about my MtDNA as well?. Just want to know if it could worth to do so in the future in my case.

Personal curiosity, let's see if somebody versed can help :P

FTDNA allows you to transfer your autosomal data to FTDNA free of charge. The 37-marker yDNA test is a once off $119 and will allow you to join as many projects as you want to. I would recommend that you first get in touch with the Y-Haplogroup I2a Project (http://www.familytreedna.com/public/I2aHapGroup) administrators before considering a move. There are a few I2a1(a) clusters that are currently being looked at for more answers.

Knovas
14-12-11, 16:24
Thanks Dorian, I'll keep it in mind if some day I have enough money.


FTDNA allows you to transfer your autosomal data to FTDNA free of charge.
¿Do you mean that my 23andme raw data would be valid to use there without any cost? Or I didn't understand this part LOL.

If that's the reality, then I supose I earn the cost derived from the raw data creation and just have to pay for the especific service, ¿right?. Also, ¿Are there other added benefits I could get transfering my 23andme raw data to them?

Thanks, and sorry to bother.

Dorianfinder
14-12-11, 16:58
¿Do you mean that my 23andme raw data would be valid to use there without any cost? Or I didn't understand this part LOL.

Family Tree DNA facilitates an upload of 23andMe raw data for $40 whereby the 180,000 (280,000 by inference) common locations between their data base participants and your 23andMe data will be compared. If you later decide to take the FTDNA Family Finder test, they will credit your $40 to that test ($199-$40= $159). 3rd Party autosomal data files may be uploaded for free and Y-DNA STR-data can be transferred from another company for $19 to have full access to all FTDNA projects.

Hope this helps

Knovas
14-12-11, 17:51
Ok, but it comes out to be more or less the same, so the main difference is the mensual billing. Anyways, I don't have money right now, perhaps one day I consider the option.

Again, thank you.

Maciamo
30-12-11, 16:44
Dorianfinder, here is a non-exhaustive list of SNP's tested by FTDNA's Family Finder (http://snpedia.com/index.php/Category:On_chip_FTDNA2) (don't forget to turn the pages, there are over 6000 of them) that potentially carry medical information. There are or surely will be thousands more, as new studies constantly find new associations between SNP's and medical conditions. It's not because an SNP has no clear medical relevance today that it won't have one tomorrow. Besides, only a small fraction of all known medical associations are found on sites like SNPedia. Companies like 23andMe have a huge stock of medical SNP's that they are leaking slowly every month to keep customer subscriptions.

Maciamo
02-01-12, 12:29
I checked a few SNP's from 23andMe health report to see if they were included on the FTDNA FF v2 chip, and many actually are. It's not just SNP's that give slight increase to have a disease, but actual genetic diseases like G6PD Deficiency are tested by the FF v1 and v2 too. The most ironic of all is that the dreadful mutation for Huntington's Disease (http://snpedia.com/index.php/Rs17781557), which is not tested by 23andMe nor deCODEme, is on the FTDNA2 chip ! As I said above, it's not because it's not in the report you get that the test doesn't carry medical information. If you don't want to know, just don't check.

yosemite
07-05-12, 12:09
Thank you both Dorianfinder and Maciamo for your debate. It helps me to ask these questions for myself as I'm considering both companies. Question for Maciamo if you please: how many markers did you test for in both Y and mtDNA tests at FTDNA?

Regarding medical genetics, it's true some diseases are caused by a single gene (and SNP), most diseases are caused by a complex gene network taking epigenetics variables into considerations. For example, diets and exercises can stop gene(s) from expressing themselves. In the Nova "Cracking your genes" (2012), Francis Collins, head of NIH, mentions he got his DNA tests from 3 companies. 23andme and Navigenics were shown in the video. In the case of type 2 diabetes, the 3 companies agree he has substantially higher risk for type 2 diabetes. But for prostate cancer, one company said he has higher than average risk, one said he has average risk, and the third said he has lower than average. What's going on? Companies often look at different parts of genes and make prediction based on incomplete information.

Cracking your genes (first aired in March 28, 2012 on PBS)
video.pbs.org/video/2215641935

Eldritch
18-05-12, 19:22
IDK what to choose between family tree dna and 23andme!

P37.2
26-01-13, 12:29
I've been actively researching genealogy for some 15+ years now and have read a lot about genetic genealogy for at least 6 years now. I came across one distant paternal cousin (according to traditional genealogy) who did a Y-DNA test with 12 markers and from that I can gather the supposed I2a haplogroup with P37.2 marker.

I've been planning to do a test for myself for some time and I am now considering mostly 23andme.com. One thing that I am not clear about is how many markers do they test for when it comes to Y-DNA haplogroup? This somehow seems to be an elusive info to find. I am interested in both autosomal DNA info as well as deep subclade info for my haplogroup and 23andme.com seems to best match these two goals at the most reasonable price.

Any input is appreciated.

Kardu
26-01-13, 17:48
At 23andme you will not get YSTR marker table to compare with your cousin or anyone else...

P37.2
26-01-13, 19:36
At 23andme you will not get YSTR marker table to compare with your cousin or anyone else...
Is this really true? I've just found this where they say there is a raw data download from 23andme.
dna-explained.com/2012/09/01/23andme-16-days-door-to-results/
I had to strip the protocol part of the link, because the board would not let me post a real link before I reach 10 posts.

P37.2
26-01-13, 19:58
Also, is it possible to take autosomal DNA data from 23andme and import them somhere else in search of closer relatives? I would like to hear from someone who has used 23andme in recent months ideally.

Jackson
26-01-13, 23:52
Also, is it possible to take autosomal DNA data from 23andme and import them somhere else in search of closer relatives? I would like to hear from someone who has used 23andme in recent months ideally.

http://gedmatch.com/

Look for the upload from 23&me section on that page.

If you are interested in relative finding or admixture, GEDmatch is the holy grail.

Alexandros
23-02-13, 15:01
Very informative discussion. Maciamo and Dorianfinder, don't take it personal guys.. It's just 2 companies.. You don't own them. Just to confirm if I get it right. From the discussion it appears that Geno 2.0 from the Genographic Project is completely redundant. Correct?

Maciamo
23-02-13, 15:38
Very informative discussion. Maciamo and Dorianfinder, don't take it personal guys.. It's just 2 companies.. You don't own them. Just to confirm if I get it right. From the discussion it appears that Geno 2.0 from the Genographic Project is completely redundant. Correct?

I don't know where you got that idea from. We have not discussed Geno 2.0 in this thread since it was only launched in July 2012, i.e. 7 months after my discussion with Dorianfinder here.

Geno 2.0 is actually the best ancestry test provided by FTDNA at present. Before that, FTDNA tests were either STR and single SNP tests, which are long, tedious and expensive, or didn't test haplogroups at all (Family Finder). Geno 2.0 is the first comprehensive test for Y-DNA that doesn't require customers to select SNP's one by one, but tests large sections of the Y chromosome (as well as autosomal DNA). In that way, it is much more similar to the 23andMe test, but is still twice more expensive than 23andMe (199.95$ against 99$) and like the Family Finder it does not report medical information. The advantage of Geno 2.0 is that it was designed specifically with the aim of improving the Y-chromosomal phylogeny.

So personally I still prefer 23andMe because I think that the medical data is at least as important as the ancestry, but Geno 2.0 has already proved useful in identifying plenty of new SNP's for the Y-DNA phylogeny.

I tend to be a bit cynical about the way FTDNA sells its products. Everything seems to be designed to extract money from gullible or overly eager customers. The 1000 Genomes Project (http://en.wikipedia.org/wiki/1000_Genomes_Project) and its follow-ups (will) provide all the full Y-chromosomal sequence (50 million SNP's), which will make the 12,316 Y-SNP's of Geno 2.0 look like a joke. It wouldn't come as a shock to me if a company like FTDNA already had access to enough full Y-chromosomal sequences and knew which SNP's could be included in the future phylogeny. How else would they select 12,316 Y-SNPs (actually half, since over 6000 were already listed by the ISOGG tree before the test was released) out of 50 million and know that exactly these would be relevant for the phylogenetic tree ? Amazing luck ? I don't believe in that.

The strategy is to make amateur genetic genealogists feel like they are the ones making the discoveries and making them pay for the privilege. When a handful of new SNP's are added to the tree each month, these become available to order for customers who original took the STR test. Many will end up paying more than 199.95$ to test all the possible SNP's for which they could be positive. I never liked that kind of game. That's almost like gambling. It may be exciting, but you always lose money. Sometimes I feel that it is because of commercial companies like FTDNA that we don't have the much-awaited 1000$ whole genome sequencing yet. We were expecting it for 2011, and it doesn't look like it will be for this year or next year either. Of course that would be terrible business for the like of FTDNA. If the WGS is only 1000$, then how much for a whole Y-chromosomal + mtDNA sequence ? Under 100$ ?

23andMe announced the imminent release of the whole exome sequence (https://www.23andme.com/exome/) for 999$ in September 2011. That is almost as good as the WGS for the purposes of population genetics or genetic genealogy. I wonder what is taking so long. Other companies like Otogenetics (http://www.otogenetics.com/human_exome_page_010812.htm) already lowered their price for the whole human exome to 698$. Too bad they don't provide the same kind of customer-friendly applications and reports as 23andMe.

Alexandros
27-02-13, 22:12
Wow Maciamo. That's a wealth of information. Almost too much to digest.. You are right, I did not notice that the previous discussion was dated before the release of Geno 2.0. Actually I am a genetic epidemiologist, so I totally agree with your views regarding the 1000 genomes data, etc. It is indeed a (very profitable) game played by these companies at the momment, as I am sure they can lower the prices even further, but decide to do so very very gradually. I am new into genetic genealogy and I find it fascinating. I have still not tested my my DNA, which makes it even more exciting for me! Actually I have motivated a number of my students (~30) to take the test with me. Unlike me, they are more interested on the medical genetics part rather than the genealogy part, therefore we will have to go with 23andme. However, I really like the Geno 2.0 brand, as the Genographic Project is the reason I initially got interested in this field (I know it is receiving a lot of criticism in this site though). I also like FTDNA's country-specific projects. I am not sure if 23andme provide a similar service and actually I wanted to ask here whether someone can enter his/her results from 23andme into an FTDNA project. For Geno 2.0 I have read that it is definitely possible.

Maciamo
27-02-13, 22:57
I also like FTDNA's country-specific projects. I am not sure if 23andme provide a similar service and actually I wanted to ask here whether someone can enter his/her results from 23andme into an FTDNA project. For Geno 2.0 I have read that it is definitely possible.

The data from 23andMe and Geno 2.0 is SNP-based, while the FTDNA projects are STR-based. The formats are therefore incompatible. I am surprised that you read that it was possible to use Geno 2.0 data for FTDNA Project. Where did you get that information ?

BakodiP
05-04-13, 14:05
The data from 23andMe and Geno 2.0 is SNP-based, while the FTDNA projects are STR-based. The formats are therefore incompatible. I am surprised that you read that it was possible to use Geno 2.0 data for FTDNA Project. Where did you get that information ?

I think it's only working when you already have a FTDNA account and tested at least the 12 basic STRs which cost now 40$. If you transfer your Geno 2.0 results to FTDNA and you're a member of a project your SNPs will appear in the Results/SNP. With these SNPs the admins of the project can sort you to a more appropriate clade then having just the 12 STRs and the shorthand.

Pi gman
06-04-13, 23:00
Is 23andme the most accurate DNA testing company?

Hi Max,
In my case I had tested with Genebase about 3 years ago and tested a couple of SNPs with them. Found out I was R1b the most common of the Y-DNA and I wanted to join projects that FTDNA offered and Genebase did not. I eventually retested with FTDNA and tested some more SNPs and found out I am R1b L2+,Z49+, and Z142+ (L20- and L196-) so far. I wish I had tested with FTDNA in the first place instead of wasting extra money. The U152 and subclades project is great especially with the maps and all.

But I may have a different goal than you. Sure, I would like to know about my Mother's side of the family but more interesting to me is my surname origin.

Also, all the testing companies are accurate but some of them test for far different allels than FTDNA and the comparisons are complicated when transferring to sites such as ysearch, etc. All in all I would say FTDNA is the place to go.

Comparing SNPs seem to be the most accurate way to find your origins and if they had the Geno 2.0 test when I first started, I would have done that one!