Page 1 of 2 12 LastLast
Results 1 to 10 of 14

Thread: Sequencing.com changes

  1. #1
    Registered Users
    Posts
    410
    Sex

    Sequencing.com changes

    Anybody tried this one out? They have new subscription model (like everybody else), though at least some people got free trial with two free app runs. (Honestly, I'm not really sure how much of this is actually applicable to new users, though it sounds like a standard marketing gimmic).

    Unfortunately, that includes conversion of genome files on the account into some sort of "genome for life" My Genomes system, which they say combines genetic data from different sources and uses "advanced statistical methods" to choose the best results out of duplicates. While I'm glad that their system appears to have been able to distinguish different samples on my account instead of creating some kind of genetic chimera, they appear to have lost some BAM files and failed to actually call variants on others.

    As that requires me to re-upload some BAM files and hope they get the processing right this time, I'm not sure if I'll be able to even avail myself of their free trial.

    There was also a promotional competition for free whole genome sequencing with the launch, which appears to have disappeared for me although it should still be running. I did notice the fine print (but really fine), once again, said only available in USA which may be why it's gone missing for me. Of course, some of their lottery entry tasks didn't really seem to work either.

    Any other observations on whether this is working or how, while I'm waiting for their support service & BAM uploads?

  2. #2
    Registered Users
    Posts
    56
    Sex
    Nationality
    Finnish
    Y-DNA (P)
    I-Y40511
    mtDNA (M)
    U5b1-a2h2d2a1a2

    Finland
    I'll try it now.. I have there already the FQs and hg19 BAM of my 30x and 650k exome genotype file. Will send 4x hg38 aligned BAM later on just to see how much it catches from it.

    Small update: After signing up it messes atleast all your previous results of EvE premium runs.. but not too surprised as sequencing.com has been pretty buggy lately (past 6 months).
    Last edited by tontsa; 11-04-2019 at 05:54 AM.

  3. #3
    Registered Users
    Posts
    56
    Sex
    Nationality
    Finnish
    Y-DNA (P)
    I-Y40511
    mtDNA (M)
    U5b1-a2h2d2a1a2

    Finland
    The "genome" is now finished.. tried for giggles SelfDecode's Cardiovascular report and it says it used "Genome1-merged.vcf.gz".. so I guess they have bunch of filters and then combine it to one vcf they can run apps on. The apprun claims there was no data on rs2106261, rs699, rs572169, rs8192678, rs1801131, rs2230806, rs1458038, rs5063, rs8192678. I'll have to manually check those from my own vcfs/genotype files to see how good that combined genome actually is.

  4. #4
    Registered Users
    Posts
    410
    Sex

    Ugh, can't even copy-paste the description on the plan selection page. "One Genome For Life" says "increasing both the quality and quantity of your data using advanced statistical algorithms". That would imply some sort of imputation, possibly with error-correction/probability. "And... One Genome automatically combines the highest quality data from each test". That's really nice, although it raises the question of how do they know "highest quality" from the DTC tests which have no quality information, or VCF files which don't use common quality base. All in all it seems like a real nice addition...

    Except that so far, it doesn't seem to really work. As reported earlier, my BAM file disappeared during the translation. I started uploading it again (Well, GRCh38 processed BAM, though it seems they will re-map and analyse it themselves...) but to test it out I also uploaded my 23andMe and AncestryDNA results. Both files created ANOTHER "My Genome" with matching name to my existing genome, but separate entry. I'm waiting to see if the status still changes, though. I guess it won't be too bad if the WGS BAM will create another, separate My Genome though.

    Prolly shouldn't speculate, but just in case... one reason for the separate genome entries could be that my original My Genome got combined with my self-imputed DNA genotypes. They might be sufficiently different from the BAM file that they just discarded the initial BAM, and from the microarray tests so they created a new genome entry. And since I uploaded the second microarray test while the first was still being processed, it may not have matched them to each other. And picking the existing genome name could just be a bug of sorts, or because they're a close match. That's all a wild guess though, and even if it worked like that, it's clear that's not a desirable outcome. Maybe I need to do some deleting & testing. Edit: Turns out deleting genomes doesn't work for me, at all. I can upload duplicate files and they just create another new genome, so it's certainly not as east as I speculated at least. They have not responded to my support requests, and since each newly created genome seems to get one bigger ID number, I can see not a huge amount are being created. This may just have been a completely premature launch from them.

    Oddly enough, now when I clicked on "Get Started" link, it tells me there's too many members and they have a queue for new members. I'm really not sure how any of this works re. old & new users.
    Last edited by Donwulff; 11-05-2019 at 10:16 AM.

  5. #5
    Registered Users
    Posts
    56
    Sex
    Nationality
    Finnish
    Y-DNA (P)
    I-Y40511
    mtDNA (M)
    U5b1-a2h2d2a1a2

    Finland
    Also as usual they don't respond to tickets. That service was suppose to include Data Viewer Ultimate.. but can't find it anywhere and running the normal Data Viewer just offers 4.99$ upgrade to "Plus". I guess I'm just going to cancel the service before they charge anything.

  6. #6
    Registered Users
    Posts
    410
    Sex

    Yeah. And BAM files just take a couple days of processing before producing... empty result files. Meanwhile, Genes for Good file they advertise supporting got stuck processing forever, but on the bright side it DID create a "Genome For Life" for README.txt in the GFG package.

    I thought I'd get at least two free analysis app runs out of it, which would be sweet as some of the apps run upwards to 200 EUR/USD (or probably more, haven't checked all) but now it seems like I can't even use that to full effect because I can't use my sequencing results. On the other hand, all of those analysis seem to be "known variants" so tests like AncestryDNA likely already cover every site they're analyzing.

    I quite understand why they're doing this change though, because as said at 200 bucks a pop few people are going to buy those analysis. A subscription creates "use it or loose it" pressure which, in theory, should significantly increase sales. I'm wondering how they're getting all app producers on board though, with both 10 bucks and 200 bucks analysis costing one use credit. The subscription is crazy low price, and the "one genome for life" has been desperately needed, but unfortunately none of the new features they're boasting seem to actually work.

  7. #7
    Registered Users
    Posts
    56
    Sex
    Nationality
    Finnish
    Y-DNA (P)
    I-Y40511
    mtDNA (M)
    U5b1-a2h2d2a1a2

    Finland
    I'm still waiting their bioinformatics people get back to me on why there are so many no calls on the combined genome. Also sucks that you can't actually download that .vcf to check yourself.

  8. #8
    Registered Users
    Posts
    410
    Sex

    The best way I have thought of for combining kits from multiple sources is to simply set every discrepant location as no-call. There is no confidence information associated with the tabulated files the DTC companies give, and sequencing BAM/VCF files are somewhat unstandardized at that point. Something like variant quality score recalibration is supposed to empirically determine them, but even that will depend on choice of "known variant" sets. You'll also have leftover variance from the sequencing platform and sample prep. (Two out of three etc. aren't actually bad, also I'm puprosefully ignoring the ability of comparing specific genotyping platform to expected variant distribution to figure which variants are "wrong")

    If you're doing that but mis-identify the reference genome (genome coordinate system) or which variants haven't been tested at all, things will go awry, which would explain missing most variants. Again a guess which is most likely wrong, just something from own experience though.

    And yeah they're really not responding to customer support, and then expect people to pay for services and support that doesn't work. Your mention of not being able to see the VCF makes me wonder if the BAM file based My Genome/Genome For Life files could actually work but just show an empty VCF file so you don't get it for free, but it's probably the simpler explanation that their analysis pipeline fails after couple of days of running and produces empty output...

  9. #9
    Registered Users
    Posts
    56
    Sex
    Nationality
    Finnish
    Y-DNA (P)
    I-Y40511
    mtDNA (M)
    U5b1-a2h2d2a1a2

    Finland
    They fixed the Data Viewer at least for me and it now finds those rsIDs.. I also tried bunch of my rare variants and they were all properly called.. compared to my own GATK best practices run. I guess I have to play more now if it actually works..

  10. #10
    Registered Users
    Posts
    410
    Sex

    Thanks for the heads up. Looked now, it looks like "Data Viewer" app now tells me it doesn't count towards my monthly uses, which may be what you mean. Still no Data Viewer Ultimate, and no response to my support requests though.

    The Data Viewer says it's loading "Samplename-merged.vcf.gz" so I guess there is a slight possibility that it's actually merging the results from different genome files behind the scenes even though they're not shown under same genome in "My Genomes", but hard to tell what's happening in the backend. I guess I need to (find) and try some variants which are discrepant in my imputed, microarray tested and sequenced variants, or Mendelian errors, but I don't remember anymore where I have all those... If they actually call variants on the BAM file despite showing 0-sized vcf, then the sequence alone should be sufficient though.

Page 1 of 2 12 LastLast

Similar Threads

  1. Sequencing.com
    By shazou in forum Other
    Replies: 61
    Last Post: 01-24-2019, 08:12 PM
  2. 23andMe sequencing
    By Flavio in forum Autosomal (auDNA)
    Replies: 1
    Last Post: 08-12-2018, 01:42 PM
  3. Oxford Nanopore's Do-it-Yourself Sequencing
    By curiousII in forum General
    Replies: 3
    Last Post: 02-12-2018, 07:59 PM
  4. What are the chemicals used in DNA sequencing
    By Anabasis in forum Inquiries Corner
    Replies: 2
    Last Post: 12-17-2015, 02:05 AM
  5. Replies: 10
    Last Post: 07-01-2015, 03:51 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •