Nebula Genomics

[aaronbee2010]

Received an email from Nebula with the offer that also said:

Already got 30x Whole Genome Sequencing? You can upgrade to 100x coverage without submitting a new sample! Contact us at [email protected] to learn more (please also share your sample ID).

I responded about the upgrade, this was their reply ...

Hey there,

Thanks for reaching out to Nebula Support, happy to help! If you previously received your 30x results, then we will need to confirm that your sample still has enough DNA to be resequenced again before upgrading.

If you have purchased within the last 30-days and your sample has not entered 30x sequencing yet, we can simply upgrade you by sequencing your data at 100x.


The upgrade fee will be $700. If you're interested please just let us know and my team will invoice you an upgrade charge.

Thanks so much,

Nebula Support

Are their any issues with resequencing a sample?

Assuming it passes QC then probably not. My Dante sample was in a vial for about 8 months before sequencing libraries were generated from it. This was from a saliva sample, whereas your results may be better due to using buccal swabs (for more on this, Trost et al. 2019 is a worthwhile read - https://jmg.bmj.com/content/jmedgene...06281.full.pdf). The level of DNA degradation your sample has undergone is probably no issue considering almost all remaining DNA will probably be much larger than the few hundred basepair lengths post-fragmentation.

If anybody here who's more knowledgeable on library prep than me could correct me if I'm wrong, that would be appreciated.

[dosas]

Back to back comparison of Big-Y700 and Nebula WGS (34x actual) on YFULL:

 

[MacUalraig]

According to the latest email re: the 100x

"you must inform us in the next 48 hours. Because we store DNA samples only for a limited time, an upgrade at a later date will require the submission of a new sample"

Is this for real I wonder.

[Brumcho]

Screenshot 2020-10-19 195949.png

I converted my Nebula cram to bam via Bam Analysys kit v2. Is it normal for my Y chromosome to have just 12x read depth? I checked my father in law's bam file stats and his Y chromosome has 10x average read depth.

[Brumcho]

[QUOTE=aaronbee2010;708777]Nebula have now started offering 100x WGS for $999 (not including membership fees) alongside their standard 30x service. Maybe pmokeefe is interested?

https://nebula.org/whole-genome-sequencing/

any advantage of doing 100x for genealogical purposes?

[aaronbee2010]

Nebula have now started offering 100x WGS for $999 (not including membership fees) alongside their standard 30x service. Maybe pmokeefe is interested?

https://nebula.org/whole-genome-sequencing/

any advantage of doing 100x for genealogical purposes?

Yes, but I don't think the advantage you get over 30x is worth the extra $700 as you'll only push a relatively small number of extra positions under the minimum threshold required for a reliable genotype determination (you're still limited by the read lengths being 150 bp). If I was at a stage where I had never purchased a WGS, the furthest I would go would be paying $499 for 50x WGS but I don't think I would go further than that. I would also be happy with 30x WGS from Nebula considering most people get over 35x mapped coverage from what I've seen.

Considering I've already purchased a 30x WGS from Dante Labs already, I would personally wait until long-read-based technologies (including linked reads) become cheaper and/or more accurate before spending money on another test for myself, especially if they're capable of resolving the ~20 million bp region of the Y chromosome currently inaccessible to existing technologies. Until then, I'm content with what I have so far.

Comparing my mothers (35x) BAM file to my maternal uncles (25x), the former had 188166/188173 loci (7 missing) on HarappaWorld and the latter got 188155/188173 loci (18 missing). I also found noticeable improvements from realigning the data with BWA-MEM (Dante Labs use DRAGEN with what appear to be suboptimal parameters - even JamesKane has observed this) so considering all of that, I don't think going all the way to 100x would yield large improvements for ancestry calculators.

As far as Y-DNA goes, Y-DNA Warehouse (also run by JamesKane) shows that Nebula 30x BAM files outperform Big-Y700 tests on average, so you probably wouldn't yield improvements with 100x WGS that justify the extra $700, considering you're still limited by the read length. You might get a small percentage of bases being resolved without ambiguity but again, nothing that justifies the extra $700.

On a somewhat related note, GATK is releasing their open-source implementation of DRAGEN next month, which hopefully has much better parameters used than Dante Labs, as they claim that DRAGEN significantly outperforms BWA-MEM in terms of accuracy. If this turns out to be true, you might be able to find a few extra SNPs from existing WGS data compared to BWA-MEM or Dante Labs' subpar DRAGEN implementation for free . I'm certainly going to be comparing GATK-DRAGEN to BWA-MEM when it comes out.

Also, before I forget - try MYTRUEANCESTRY10 on Nebula and see if it nets a $30 discount. I just tried it and it worked for me. Just putting that out there for anyone interested. It only appears to work for the 30x test.

[TigerMW]

Back to back comparison of Big-Y700 and Nebula WGS (34x actual) on YFULL:

It is a bad idea to use YFull statistics to compare between tests. First and foremost, they don’t realign the BAM files. They told me this when I asked. That means their comparisons are apples to oranges. This can also cause problems in phylogenetic analysis.
The Y DNA Data Warehouse does realign the BAM files using one method (they say and I believe).

[aaronbee2010]

Screenshot 2020-10-19 195949.png

I converted my Nebula cram to bam via Bam Analysys kit v2. Is it normal for my Y chromosome to have just 12x read depth? I checked my father in law's bam file stats and his Y chromosome has 10x average read depth.

WGSExtract considers all 57227415 bp of the hg38 chrY reference when calculating coverage, even though most of the chrY reference has no reads able to cover it, and these regions need to be ignored when calculating average coverage. For reference, I get 10x chrY coverage on WGSExtract and 13.37x on YFull after they apply their filters. dosas himself got 18.51x coverage after filtering, so you and your father-in-law should get statistics much closer to dosas YFull coverage than mine, assuming microbial contamination is not excessive.

[aaronbee2010]

It is a bad idea to use YFull statistics to compare between tests. First and foremost, they don’t realign the BAM files. They told me this when I asked. That means their comparisons are apples to oranges. This can also cause problems in phylogenetic analysis.
The Y DNA Data Warehouse does realign the BAM files using one method (they say and I believe).

It says in the picture that both the BY700 and WGS tests are aligned to hg38 already, or am I missing something?

[dosas]

It is a bad idea to use YFull statistics to compare between tests. First and foremost, they don’t realign the BAM files. They told me this when I asked. That means their comparisons are apples to oranges. This can also cause problems in phylogenetic analysis.
The Y DNA Data Warehouse does realign the BAM files using one method (they say and I believe).

What apples and oranges. You give them the .CRAM and they align it themselves. Your info is wrong and/or outdated. You can see it on the screenshot both files are aligned.