The inconsistencies of the BigY test - How to Manage?

[djknox]

It has come to light after reviews of my BAM file that I definitely possess certain "novel variants" that for whatever reason FTDNA did not pick up on or report in my BigY interface. The exact reasons for this I am still trying to understand... but anyone involved in BigY will know that many SNPs are not black & white.

My question is, knowing this and uncovering new data manually using 3rd party tools, how can I keep my BigY interface current and up to date? If I send a review request to FTDNA, will someone look at and correct a specific SNP report in my BigY interface?

Any thoughts?

[wighty44]

I have 2 SNPs that were not identified as negative on my Big Y Tree yet were clearly shown negative after YFull analyzed my BAM file - L366 & CTS11142.

I wrote to FTDNA's help desk asking for them to have my account's Y haplotree updated to show that SNPs L366 and CTS11142 tested negative. I even sen FTDNA YFull's SNP results data parameters among which showed a 0.0 probability of error.

After 10 days without any reply (other than the immediate automated reply), I sent another email to which FTDNA provided this response,

"Thank you for your email and I apologize for the delayed response. Concerning kit number XXXXXX, the Big Y results show negative for CTS11142. I see that this is not listed. I went ahead and forwarded this issue to our IT team. Concerning L366, this marker is a deletion which does not show up in the Big Y FTDNA results but can in the BAM file."

The last time I checked my Y Haplotree had not yet been updated. I tried to check just now , but it seems FTDNA is experiencing website problems because the haplotree page fails to completely load, and the main page is displaying in an odd format.

[lgmayka]

My question is, knowing this and uncovering new data manually using 3rd party tools, how can I keep my BigY interface current and up to date? If I send a review request to FTDNA, will someone look at and correct a specific SNP report in my BigY interface?

You are not describing a bug in your own personal account, but rather a poor design decision that has affected all Big Y results since mid-June.

My best guess is that someone decided that for privacy reasons, private SNPs would not be presented to project administrators. The problem right now, of course, is that they aren't presented to the customer himself either, except in the BAM file and perhaps in the VCF file.

[djknox]

Yes agreed that this isn't an isolated issue unique to my test. But if the tools offered via the BigY interface are to be useful, there must be a way to update data that is clearly incorrect or missing. I understand that some of the data is very open to interpretation... but such expensive and detailed tests should not be offered unless there is a mechanism to keep the data relevant.

The more I read comments about this stuff, the more it is very evident that tests are fraught with inconsistencies & read errors along with a lot of interpretation issues.

Although I must disclose that this is a common theme with my perspective on genetic genealogy, I really feel consumer products must pass muster at a reasonable level. As it stands, only scientists schooled in genetics can make much sense of these testing/reporting issues.

I don't expect the test and its report to be flawless, but equally I don't expect that a casual stroll through these BAM files can find so many problems.

I wonder if the reporting of tests already completed will improve with time... or are we stuck with what was given to us at the getgo?

[lgmayka]

I don't expect the test and its report to be flawless, but equally I don't expect that a casual stroll through these BAM files can find so many problems.

An immediate step that FTDNA could take is to make a deal with YFull or FullGenomes, such that each Big Y customer would get a coupon for free or half-price professional analysis.

The longer-term alternative is for FTDNA to either buy one of those two companies, or reproduce the expertise in-house (which would probably mean re-hiring the Krahns).

I wonder if the reporting of tests already completed will improve with time... or are we stuck with what was given to us at the getgo?

My suspicion is that FTDNA will progressively improve the presentation of data from the VCF and BED files, but will not reprocess the raw data--unless FTDNA actually does buy or reproduce the expertise of YFull or FullGenomes.

[djknox]

Interesting Igmayka. Are you suggesting that the other 2nd generation testing companies are offering much better service? Up until now I try to keep all the tests in one website just to keep management efforts to a dull roar. I had tried 23andMe and Ancestry and it is very obvious their interests are different from mine.

Right now I just don't have the time to spend the many hours needed to get up to speed at the level that Big Y demands for self analysis. Group administrators can only go so far (and of course they themselves have many different levels of analysis capabilities).

So, FTDNA, what's the plan for administering BigY on an individual basis?? At least their should be a paid service to deeply interpret and correct results? There must be some hungry genetics students somewhere looking for some part-time income... why not set up a co-op with a university somewhere?

[djknox]

I just noticed that YFULL IS an analysis service. Anyone have experience with them? $49 seems like a bargain if in fact they can clean up by Big Y results? I kind suspect however that they just offer some automated output... which puts me back to where I began...

[wighty44]

I support YFull's analysis service

I just noticed that YFULL IS an analysis service. Anyone have experience with them? $49 seems like a bargain if in fact they can clean up by Big Y results? I kind suspect however that they just offer some automated output... which puts me back to where I began...

I'm not sure what your specific testing goals might be, but I found YFull's analysis results very clear, well-organized, detailed, and helpful.

Helpful in three ways. First, it identified two R1a SNPs as negative, that FTDNA still shows as "available" for testing on their version of my Y haplotree. After paying hundreds of $$ for the Big-Y test, why should it not provide complete SNP results, or mislead me towards unnecessarily spending more $ on SNP tests whose results are negative in my Big-Y BAM file?

Secondly, YFull's hapoltree is more responsive and reflective of newer branches.

Thirdly, YFull has been more informative with monthly updates that identify new Y-tree changes and new samples that have been added to various tree subclades. I have never received such notices from FTDNA. I find FTDNA's customer service MIA.

IMO, I feel that YFull fills a customer need absent from FTDNA as it seems it was designed with the customer in mind, and I'm very glad it exists...

[Pinny]

I'm not sure what your specific testing goals might be, but I found YFull's analysis results very clear, well-organized, detailed, and helpful.

Helpful in three ways. First, it identified two R1a SNPs as negative, that FTDNA still shows as "available" for testing on their version of my Y haplotree. After paying hundreds of $$ for the Big-Y test, why should it not provide complete SNP results, or mislead me towards unnecessarily spending more $ on SNP tests whose results are negative in my Big-Y BAM file?

Secondly, YFull's hapoltree is more responsive and reflective of newer branches.

Thirdly, YFull has been more informative with monthly updates that identify new Y-tree changes and new samples that have been added to various tree subclades. I have never received such notices from FTDNA. I find FTDNA's customer service MIA.

IMO, I feel that YFull fills a customer need absent from FTDNA as it seems it was designed with the customer in mind, and I'm very glad it exists...

Just remember that the YFull's Experimental YTree is just that Experimental. One thing that I remember reading is that some of positions of the public kits ie. 1KG results are incorrectly placed throughout the tree.

I really like the site and their analysis of the results but just remember it is an interpretation of the results. Until the results are confirmed against several testers they may be misplaced on any tree.

[lgmayka]

One thing that I remember reading is that some of positions of the public kits ie. 1KG results are incorrectly placed throughout the tree.

Do you have any specifics on this claim? If so, I'm sure that YFull would be glad to hear them.

But keep in mind that early public genome scans (e.g., 1000 Genomes project results) are of very low quality, and therefore subject to plenty of interpretation. I could imagine YFull classifying them either more or less conservatively than others might, especially if no customer has submitted a high-quality sample in the same phylogenetic neighborhood for comparison.

I really like the site and their analysis of the results but just remember it is an interpretation of the results. Until the results are confirmed against several testers they may be misplaced on any tree.

YFull waits for two samples, of course, to define a new subclade; but I agree that further samples may prove the first construction wrong--e.g., if a SNP thought to reliably mark a subclade turns out to be unreliable.

I would go farther and say, with ISOGG, that any haplotree based solely on Next Generation Sequencing (NGS) is tentative until verified with another technology. That is why ISOGG specifies that new subclades defined through NGS alone must appear in italics.

[lgmayka]

$49 seems like a bargain if in fact they can clean up by Big Y results? I kind suspect however that they just offer some automated output.

Yes, it is well worthwhile, if you have patience. (Full analysis may take a month or so after submitting your BAM file.)

No, YFull does not merely emit a set of files, but rather provides an interactive interface (with the additional capability of generating CSV files). YFull provides:
- Classification and placement on their frequently updated haplotree
- A list and count of all your named/shared SNPs (positive, ambiguous, and no-call)
- A list and count of all your unnamed/private SNPs, classified according to quality (best, acceptable, ambiguous, low, one-reading, indel)
- Over 400 Y-STR (marker) results, including about 100 of FTDNA's standard 111
- The ability to look up your precise result counts (e.g., "56 G, 1C") at any named SNP or numeric location
- Almost complete sequencing of your mtDNA, with subhaplogroup classification
- Ability to join a group/project, to facilitate comparisons.

To be fair, I must point out that FullGenomes also offers professional analysis for $50. As far as I know, however, they generate text files.

A third service, clarifYDNA, "merely" generates a haplotree report from your VCF/BED files, for $30. Its primary advantage is promptness:
- It uses VCF/BED files (which the customer can download himself) rather than the BAM file (which FTDNA must supply upon request)
- It reportedly returns a result within a week.

[Pinny]

Do you have any specifics on this claim? If so, I'm sure that YFull would be glad to hear them.

I have attempted to find an email address for them but there isn't one listed on their website.

[djknox]

lgmayka you touch on the heart of my issue. These 2nd generation tests are LOW RESOLUTION SCANS designed to throw a wide net. OK Great. But where do we go from here? I have dozens of pertinent SNPs that are negative when they should be positive... based on group testing and a cursory understanding of my BAM file. So should I now be trying to order individual SNP tests to confirm or reverse BigY results... noting that most of them (ie the SNPs) aren't likely offered yet.

I get the feeling that all I can do is either become a geneticist... or wait for Others to sort things out.

For those who know YFULL - how do they provide a customized service for such a low price? I am a little suspicious.

[tlgarrison]

For those who know YFULL - how do they provide a customized service for such a low price? I am a little suspicious.

PERHAPS YFULL is more about the research and the $49 is just what it takes to keep it going without pricing them out of a BigY donor? Research costs money but if they ask for more than $49 BigY and other donors would start to drop out (less kits to use for research). If they are true researchers then they would want to maximize the number of participants.

I belong to r1a1a (A3) project and won't get my BigY until end of September but I will use the YFull service and the help of the r1a1a admins to hopefully move me beyond my (tested) Z282 designation.

[lgmayka]

I have dozens of pertinent SNPs that are negative when they should be positive.

Do you mean that your Big Y results assert definitively that you are negative for a "known" (named) SNP when you know you should be positive?

Or do you mean simply that a "novel variant" does not show up in your list, and you think it should?

For those who know YFULL - how do they provide a customized service for such a low price? I am a little suspicious.

To be fair: Much of the analysis appears to be automated. Only the most crucial part--deciding which SNPs and SNP results are the most reliable--is at least partly a human judgment (and therefore also potentially subject to revision based on later evidence).