Clare86
Has anyone ordered and received an upgrade to GenomeH?

If you did how long did it take and what did you get?

I ordered 12 weeks ago and I still haven't received it, apparently they can't give me an estimation of when I will get my results.
Quote 0 0
Erik Chapman
I'm having the same issue from the same time period.  I'd speculate that they got more than they expected with that offer and now have a backlog, but they aren't responding to clarify or give a time frame.  That they've taken down the pages offering long reads or hybrid genome is disquieting, it makes me wonder if there's some fundamental glitch with that machine, or worse, that the greater presumed accuracy with the longer reads has exposed some inaccuracy in the short reads that they are suppressing. 

It also isn't clear to me why they haven't rolled out the exome x130 component of the hybrid, given ordering that alone would have been three weeks and would give at least something to work with.   They've also stopped updating my complete reports package, although I'm paying to have it updated monthly.  

I've had various illnesses for decades where I suspect a genetic origin and I was hoping to narrow the scope of what to look at by getting rid of some false positives.  Even with an inaccuracy chance of 1 in 10,000, with hundreds of thousands of base pairs in the literature now that can be a lot of false positives.   Also for local help with say blood testing for the accuracy of some possibilities, I was hoping to have closer to diagnostic levels of certainty on the genetic end to get doctors interested.  

One thing I'm doing in case I'm waiting a while for the H test is ordering from Nebula, which will give me a chance to check for consistency and potentially to combine for greater accuracy if I can figure out how to do it with something called Samtools. 
Quote 0 0
Clare86
I'm sort of glad I'm not the only one but also sad there's another person being irritated by Dante over this.

I had already upgraded previously to x130 so I am only waiting on the long reads. In one of the emails I have had from them over this they also said that they weren't actually going to be doing a hybrid assembly and are just going to provide fastq files for the long reads and are not going to make bam/vcf files for the long reads. I complained about this as that's not the product they described. I included Andrea in my complaint email and he said they would provide a "custom solution" for me, whatever that means.

I particularly wanted the long reads as I'm interested in learning about any copy number variants I might have. I have Asperger's and CNVs are quite common in autism, so I am quite disappointed to not have this yet.

I've gotten Nebula's x30 sequencing too as well as Exome sequencing from Sano Genetics. The data from both looks pretty good, I've been able to confirm all the interesting variants I found in my Dante data.
Quote 0 0
Erik Chapman
That definitely wasn't what they promised.  I'm attaching a screenshot of the offer from my email.  They promised the hybrid information would be used for *everything*.   screenshot Dante upgrade.png 
Quote 0 0
Clare86
Yeah, that's why I was really surprised when they told me that's not what they were going to provide. 

I get that sometimes things happen and companies can't provide the exact product they offered but Dante really should be providing clarity on what problems they are having, what impact those problems will have on the product and if they can't offer the described product, offer something of equivalent value or a refund.

Personally I think if they are not going to provide a hybrid assembly then at the very least they should do CNV variant calling on the long reads. Not sure about snp/indel variant calling as that doesn't tend to be very accurate on just long read data. Most people wouldn't be able to do this themselves as from what I can tell it requires the download of extremely large files.
Quote 0 0
Erik Chapman
It would require additional statistical analysis and programming.  The statistical analysis could potentially result in a lot more accuracy from combining the two methods, but they'd have to do the work so that each could help interpret the other.  That could include reinterpretation of existing data to a degree with the influence of the long read data.   The results for long read would presumably be somewhat outside the ranges the existing programming for producing short read fastq so there would have to be programming for the translation to put them together, or more likely a new type of fastq which could consistently incorporate the data from both.  Having one translated into the other would somewhat defeat the purpose of fastq being the original data, so a format with the original data of both is needed.  That might produce issues with third parties and software using the blended data though. 

There should actually be less of an issue with figuring out the blended consequences such as BAM given that they are already derivative.  Take BAM files, they involve the call and the level of certainty, number of reads etc.  Producing one that accurately reflects those conclusions isn't conceptually that difficult, but it would require additional programming specifically for these hybrid cases, which might be a small proportion of their overall orders. 

That said, they shouldn't have offered it until they were ready to do it.   I asked them specifically about whether some of the restrictions in place for Hybrid before the offer were still there, and they said no, everything would be as before but with more information.  
Quote 0 0
Clare86
I have just noticed that Dante are no longer referencing GenomeH on their coverage upgrade page. Not sure what that means for us though.
https://www.dantelabs.com/products/whole-genome-upgrade
Quote 0 0
Erik Chapman
It's hard to say, it's equally consistent with a number of different theories.   Their offerings and prices have tended to vary wildly in a way that is probably driven by demand.   The high base prices now suggest that demand is high.   Key question, what is being done with the resources used to make the long reads?  Did the machine break down or get repossessed or are they too far behind on orders?  They could also have had a fundamental flaw in some underlying analysis for blending the two types of reads.  

Early on I remember the site used to say that the reports weren't available for either the long read genome or the hybrid, that they were more for research purposes.  Then rather abruptly with the half off offer, they claimed that the upgrade could be used for everything.  I asked about that issue specifically and was told it would be just the same, but with a lot more information.  I couldn't find the reference to the hybrid not being eligible for the reports so they appeared to have removed that.  

Two possibilities to do with development could be involved.   One is, they might have thought they were close to a breakthrough on developing a method to integrate the two types of reads and a glaring flaw has emerged, another is that they anticipated completing development but got sidetracked.   The philosophy of Dante so far has seemed to be that it is better to try to do twice as much and miss half the time than to reliably do exactly what was promised.   

So they have at least a miss with developing the hybrid/long read business, which is central to their strategy and goes to the uniqueness of their brand, while at the same time racing to get out yet another marginal me-too Covid product just like every other bio company.  They also raced to get out a bunch of new reports, before correcting the deficiencies in the prior ones, and apparently at the expense of not keeping their promise to update reports for subscribers.   I bought the all panels option with a subscription but only 2 panels have been updated in the past 3 months.  

They have a real issue with not finishing things before chasing the next shiny object distraction. 

The longer they don't answer me, the more it looks like there was some fatal flaw in their whole idea.   In which case, it would have been better if they'd owned it and dealt with it.  They've already lost orders I've sent elsewhere because I can't rely on them.  If they'd owned a problem and offered a refund without playing games, they probably would have got those orders.  That was an option I was thinking about a month ago.   But I don't want my spouse and my dad to have to deal with these guys while they're this evasive. 

The irony is, what I'm really after is precision, and for most purposes I could probably get a lot more precision with a lot more standard reads.  I went for the hybrid option because it presented the likely highest degree of precision they were offering for a regular product.  It was really proportionately expensive though.  Adding in 15x long reads for the price of 3-4 30x genomes.   My first I got an average of 39x, if I could get 150x+ across the board in a format that would be usable with standard programs [not sure if there is a cap on usable fastq] and with the vcfs, BAM etc based off the whole thing, I'd probably prefer the result to 15x long reads.   Especially if I could get a couple of custom things that will be vexing for me to figure out myself, such as a full unfiltered VCF.  


  
Quote 1 0
Clare86
I sent Dante an email asking for an update on Thursday and yesterday afternoon they replied saying that are cancelling the GenomeH upgrade and have refunded my money. They said that "due to the problems with a provider caused by the covid pandemic we are unable to provide the long reads sequencing needed to complete the WGH upgrade".
Quote 0 0
Erik Chapman
I hope they at least give me the 130x exome at pro-rata price, since it was a component of the 50% off deal.  I'm going to get stung twice on the exchange rate if there's just a refund in Euros to the credit card, in each direction.   I'll be extremely annoyed if they want me to pay full price and take a third exchange rate hit on substantially the same transaction.  
Quote 0 0