Monday, 20 January 2014

HiSeq NextSeq 500

Onto the next new instrument from Illumina, the NextSeq 500, I've given a brief overview but covered the new chemistry in an earlier post.

Illumina helpfully provide the NextSeq 500 User guide on their website. There is a lot of interesting reading in it and I’ve picked out a few of my favourite bits. The datasheet also describes “streamlined Illumina sample preparation kits”, I’ve not looked to see what changes have been made yet but Keith Robison over at Omics Omics has a post which describes the new NeoPrep system from Illumina. This is Illumina's incarnation of the Advanced Liquid Logic technology they purchased last year; expect 16 DNA or RNA library preps per run using TruSeq and/or Nextera assays.

Running NextSeq 500: NextSeq 500 borrows the concept of high-output runs from 2500 and the micro & nano-kits from MiSeq, allowing users to choose from High-, mid- and low-output modes. The datasheet does not mention the low-output mode but the user guide suggests 20Gb or around 60M clusters per flowcell, mid gives 40b and 130M and high-output is 20Gb and 400M reads. Single-use reagent kits contain: the flow cell and clustering & sequencing reagents. Flowcells contain four lanes but only two inlet ports and a shared library space in the reagent cartridge. This means similar to MiSeq and HiSeq 2500 rapid-run mode only a single library or library pool can be sequenced across all four lanes.

I think that SuperPooling, i.e. pools of pools, is the future and this seem to confirm Illumina's expectation of how users want to be sequencing. No problem for users who need all the data for a single library, and not much of a problem for users mixing the same library type where pools have compatible indexes. However try mixing a cocktail like Nextera and smallRNA libraries and achieving 50:50 and things are likely to go horribly wrong.

New Chemistry: NextSeq 500 uses a two-colour chemistry (described here) rather than the original four-colours. This makes a massive difference to the complexity of producing reagents, the instrumentation and the computation; all are effectively reduced by a factor of two. The simpler, and therefore cheaper cameras mean Illumina put six into NextSeq 500 and can therefore image both lanes at the same time. Cameras 1–3 image lane 1 while cameras 4–6 image lane 3. Then the imaging module moves on the X-axis to image lanes 2 and 4.

A NextSeq 500 flowcell
Speed: The NextSeq 500 is now Illumina's fastest "genome" sequencer running at a maximum of 10.34bph (base-pairs per hour) compared to HiSeq 2000's 1.73bph and HiSeq X Ten's 4.16bph. Proton still kicks ass on paper as far as speed goes with 50bph, but this may turn out to be the only metric LifeTechnologies can point to in marketing literature without too much spin.

So should you buy one: It will be interesting to see how users respond to the speed of NextSeq 500. I think it may well become the default clinical sequencer as it offers a price-point labs can probably afford, offers flexibility in per run pooling via different flowcell configurations and is fast. Will current HiSeq 2500 rapid run users really be tempted to throw out one box and buy a new one remains to be seen. I think 2500 is here for a couple more years, I certainly hope so as we just bought one!

1 comment:

  1. Geneticist from the East21 January 2014 06:04

    Hi James

    Can you clarify this?

    "HiSeq 2500 rapid-run mode only a single library or library pool can be sequenced across all four lanes."

    As far as I know, you can load two different libraries on the two Rapid lanes using the Truseq Duo cBot Loading Kit. I am wondering if similar kit can be made available to NextSeq 500 as well.