Opiniomics

bioinformatics, genomes, biology etc. "I don't mean to sound angry and cynical, but I am, so that's how it comes across"

Illumina destroy the opposition again – almost….

Illumina today announced two new systems, the NextSeq 500 and the HiSeq X Ten.  I think we can all see that the NextSeq 500 is there to compete with (a.k.a. destroy) the Ion Proton, and in their own words, the HiSeq X Ten system is there to “sequence entire nations”.

From this link here, I have created the following table (I implied the number of lanes per flow cell to make the numbers match up, I assume they are correct, but please say if not):

Reads per flow cell Read Length Pairs Flowcells Gb Tb
MiSeq

25000000

300

2

1

15

0.015

2500 Rapid

300000000

150

2

2

180

0.18

2500 High Output

2000000000

125

2

2

1000

1

NextSeq Mid

130000000

150

2

1

39

0.039

NextSeq High

400000000

150

2

1

120

0.12

HiSeq X

3000000000

150

2

1

1800

1.8

HiSeq X Ten

3000000000

150

2

10

18000

18

These are pretty amazing announcements from Illumina, but I want to get on to the politics of this rather than raw statistics.

NextSeq 500

The NextSeq 500 will retail for $250,000 apparently, and this will really make people think when previously they may have bought a MiSeq or a HiSeq 2500 (or an Ion Proton).  It’s a really interesting entry into the market and will shake things up as smaller labs become capable of really high throughput sequencing.  120Gb is 1 human genomee at 40X coverage, and 400 million reads is 16 RNA-Seq samples at 25M reads per sample.

Really, though, this blows the Ion Proton out of the water.

The HiSeq X Ten

This is the real juice of the announcment.  Here are some additional facts:

  • You cannot buy 1 HiSeq X instrument.  You have to buy 10 (the HiSeq X Ten system)
  • It is limited to human only

This is clearly an interesting move.  I chatted with an unnamed source from Illumina and they told me a few things.  The “human only” mode is nothing to do with chemistry, it is to do with software, which means you can probably hack it to sequence other species, but you’ll likely break your warranty.  In my view, this is clearly aimed at competing with Complete Genomics.  I was at ICG8 recently and CG claimed a 30X increase in throughput.  Clearly Ilumina had to respond and the HiSeq X Ten is it.  CG is limited to human, and as soon as CG compete in other species (which they will), Illumina will too.  The big difference is that CG’s limitation is technical, whereas Illumina’s is imposed by software, and I don’t know how the non-human communities are going to take that.

The other interesting thing I gleaned, and I quote, is that Illumina are “being approached by whole countries asking us to sequence their populations”.  Clearly the HiSeq X Ten is the system they will use.

HiSeq X Ten costs

The X Ten claimes the $1000 genome, and this was broken down as $800 on reagents, $135 on equipment depreciation over 4 years, and $65 on staff/overheads.  Now, from the datasheet, the X Ten system claims 18,000 human genomes (30X) per year:

HiSeqX10

So let’s do the maths: 18,000 genomes * $135 depreciation per genome * 4 years = $9.72 million, which squares with the rumour of  $10M price tag for the X Ten system.

18,000 genomes * $65 = $1.17 million on staff and overheads to run the thing per year.  At many institutions I know of this would be enough to employ ten members of staff.

So actually, I think they might be right in claiming the $1000 genome – if you do 18,000 human genomes per year for four years on each X Ten system.  That’s a lot of human genomes though…..

On Bioinformatics

The obvious question is “does the $1000 include bioinformatics?” – well, with the recent release of the ISAAC aligner, and developments in BaseSpace, I think the systems will go all the way to SNP calls at little or no extra cost.  Is that bioinformatics though?  The interpretation (certainly in humans) will need to be done by clinicians, and no, the cost of the clinician is not included in the $1000 😉

The Competition

With the NextSeq 500 they are competing with Ion Proton, and with the X Ten they are cornering the human market and competing with CG.  I say competing, but actually this is more of a slaughter.

The elephant in the room is Oxford Nanopore, and the fact Illumina are focusing on Ion Proton and CG seems to suggest they don’t think that ONT are a threat.

With the MAP recently announced, they may be wrong on this – only time will tell!

21 Comments

  1. Nice post. Wondering whether the number lanes in the table for NextSeq are right. It looks like NextSeq High-Output and Mid Output flowcells have four lanes (fat and slim). Not sure, but check the pictures Illumina main site in the tweets below.

    A first look at @Illumina NextSeq 500 High-output flow cell (and 4 fat lanes?). #NextSeq pic.twitter.com/pSjqehr8cS— nextgenseek (@nextgenseek) January 15, 2014

    A first look at @Illumina NextSeq 500 Mid-output flow cell (and 4 slim lanes?). #NextSeq pic.twitter.com/pytQtSuY9o— nextgenseek (@nextgenseek) January 15, 2014

  2. They’re not right, but the table is ripped from the illumina website b where they have a “reads per flowcell” and “total output” figure, so I implied the number of lanes. The total output figures are correct.

  3. I have removed all mention of lanes, I am too jet-lagged!

  4. Exciting as you say. Personally for me the MiSeq pricing is the most exciting, this will let everyone get in on the action, particularly if directly translated to UKP (£60k!)

    Correction: 120Gb is actually one human genome at 40x coverage.

  5. Thanks a lot. Take rest and write more posts 🙂 I am really intrigued by the new flow cell designs.

  6. Looking over the data sheet for the HiSeq X Ten (http://res.illumina.com/documents/products/datasheets/datasheet-hiseq-x-ten.pdf) under system specifications, glad to know illumina have made this clear: “For Indoor Use Only”

  7. Thanks Nick, I shouldn’t be doing maths when I am jet-lagged! (or tipsy, Illumina gave out free beers yesterday)

  8. I might be wrong but I think I read somewhere in their literature that each HiSeq X machine has 2 flowcells

  9. Just read the datasheet… the quoted 1.8 TB run is on a DualFlow cell configuration

  10. Mick, I think the HiSeq X is targeted at destroying the exome market completely, with a particular eye on Agilent, but also throwing all other enrichment-based genomic content concentration assays under the bus along with them. By making a system that can sequence whole human genomes at $1,000 but that they’ve artificially locked out of performing exome and other enrichments, they’re basically making people choose between a whole human genome and an exome. Off the top of my head, the per-base cost of sequencing alone on any enrichment is going to be at least 2.5x as much.

    People being able to enrich exomes cuts into their margins on sequencing by allowing higher sample throughput, and they were unable to take over the enrichment market like they did with sequencing, so they want to leapfrog it completely with the HiSeq X. They know many people will choose to go for a whole genome rather than exome if the price is nearly the same, and they can ensure the price stays very close because they dictate the reagent costs on both systems (thanks to a lack of competition in the sequencing space).

    I believe this was the major impetus, because it’s the only real area where Illumina has competition with Agilent being the other major player in the exome enrichment space now.

    At first blush, this may seem like a boon for researchers who always want “more data”, but there are potential problems, I think.

    One potential scientific problem with it is that it’s a whole human genome sequencing run as defined by Illumina. I think there’s an open question as to whether Illumina’s choice of how to define what constitutes a “whole human genome” run will mesh with individual researchers’ definitions. Another potential problem I see with it is that it further concentrates human sequencing in the few centers that can afford the $10M buy-in, which I don’t see as a good thing because it could further reduce competition in the services space.

    Anyway, those are just the issues with it that pop to mind. On the one hand, I love the idea of the $1,000 genome. On the other, I have some reservations about it.

  11. Yes, I think it’s been clear for some time that Illumina would prefer us all to be sequencing genomes and not exomes, but I disagree that this is the major focus of the X Ten – I think it’s more of an added bonus. The X Ten is an attempt to out-compete Complete Genomics, in my opinion.

  12. Yeah, I think the opposite. Basically, I think their goal is to replace exomes with whole genomes and being able to put the final nail in the CG coffin is a happy side effect.

  13. I suppose it depends on who illumina was losing more money to, Agilent or Complete Genomics? I think it should be viewed as a two-birds/one-stone stroke. And no one is talking about the massive, massive amount of data that these things are going to generate. $10M to by an X10, another couple million to pay for enough storage?

  14. So the total operation expenditure to get the 1000 dollar genome totals 72 million over 4 years huh?

  15. I like this great advancement made by Illumina, but … 150bp for whole genome sequencing? You better keep to re-sequencing with a known reference genome, man.

  16. Genohub.com is looking to match pent up WGS demand (wait times for access to larger WGS facilities is still significant) with new HiSeq X Ten supply: http://blog.genohub.com/nextseq-500-and-hiseq-x-ten-services-coming-soon-to-genohub-com/

  17. “Keeping up with a $10 million HiSeq X 10 system may require a compute system costing on the order of $450,000 in capital up front” : http://glennklockwood.blogspot.com/2014/01/the-1000-genome-computational.html?spref=tw

  18. Here’s the way I view the whole thing. I bought illumina stock in September. it’s done very nicely. but you know what I am going to buy now? PacBio.

  19. The NextSeq 500 is effectively single lane from the users perspective. Running multiple samples per NextSeq run will be done via multiplexing.

  20. Macrogen mentions in its website that they will accept samples for “X genome” service:
    http://www.macrogen.com/eng/business/xgenome_product.html
    Can we have genohub updated to include this service when the details are announce?

  21. We’d certainly like for Macrogen to list their X Genome Service on Genohub.com. You can get started here: https://genohub.com/ngs-providers/

42 Pingbacks

  1. Illumina destroy the opposition again - almost....
  2. Illumina announces new sequencers - NextSeq 500 and HiSeq X
  3. The $1,000 Genome Arrives -- For Real, This Time - Forbes
  4. Blog David Singui » Science enters $1,000 genome era
  5. Henrik Maxian | Science enters $1,000 genome era
  6. Science enters $1,000 genome era | PDF Yann Ketchanga
  7. Pour mille balles, t’as un génome humain | Tout se passe comme si
  8. Pacific Piano Performance Posts | Science enters $1,000 genome era
  9. Science enters $1,000 genome era | Today News Time
  10. D Richard Dance Blog » Blog Archive » Science enters $1,000 genome era
  11. Science enters $1,000 genome era | RizkiPurnama.Net
  12. Science enters $1,000 genome era | NHAT NET
  13. Science enters $1000 genome era – BBC News | Best Food For Weight Loss
  14. Science enters $1000 genome era – BBC News | Healthy Eating
  15. Science enters $1000 genome era – BBC News | Everyday News Update
  16. Science enters $1,000 genome era | EikAwaz.com
  17. Science enters $1,000 genome era | RocketNews
  18. Science enters $1000 genome era – BBC News | Latest News Portal Info
  19. Science enters $1000 genome era – BBC News | Unexplained Weight Loss
  20. 7 reasons why the NextSeq 500 is a strange choice | opiniomics
  21. The $1000 Genome? – Not Exactly | DNAeXplained – Genetic Genealogy
  22. NextGenSeek Stories To Read This Week (1/17/2014): Illumina, HiSeq X, NextSeq 500, Roadmaps
  23. Human genomes can be sequenced for $1000 – DigitalJournal.com | Best Food For Weight Loss
  24. Science enters $1,000 genome era | RizkiPurnama.Com
  25. Science enters $1,000 genome era | D Richard Dance Website
  26. Human genomes can be sequenced for $1000 | NHAT NET
  27. Science enters $1,000 genome era | Twana Sparks Blog 2
  28. Henrik Maxian Blog - Science enters $1,000 genome era
  29. Science enters $1,000 genome era | Dr Twana L Sparks
  30. Yann Ketchanga data | Science enters $1,000 genome era
  31. Science enters $1,000 genome era - The Cure within One Minute
  32. Links 1/20/14 | Mike the Mad Biologist
  33. Illumina introduces world’s first $1,000 genome » Oxbridge Biotech
  34. The de-democratization of sequencing | opiniomics
  35. Sequencing Has Always Been Undemocratic | Mike the Mad Biologist
  36. On cheer-leading and pom-poms | opiniomics
  37. The $1000 genome, what does it mean for patients? | The Family History of Bowel Cancer Clinic
  38. Biome | $1,000 genome
  39. Welcome to the $1000 genome | opiniomics
  40. Time for New DNA Synthesis and Sequencing Cost Curves - By Rob Carlson - Synbiobeta
  41. Edinburgh and Glasgow invest in revolutionary HiSeq X clinical genomics technology | opiniomics
  42. Welcome to the $1,000 genome: on Illumina and next-gen sequencing - Biome

Leave a Reply

© 2017 Opiniomics

Theme by Anders NorenUp ↑