Note: The NovaSeq X+ prices listed further down this page have changed since this was first published; you can find up-to-date pricing information on the DNA Sequencing core’s website. The relevant price would be listed under “10B 2×150 Full Lane”.
The UW Biotechnology Center DNA Sequencing Facility has begun rolling their new NovaSeq X+ instrument into operation as a replacement for the NovaSeq 6000. This change will significantly impact the cost of sequencing for single cell RNA-seq projects at UWBC, lowering one of the major barriers to entry for first time users and enabling researchers to plan larger, more powerful studies on the same budget. In this post, we will discuss some of the key differences between the X+ and the 6000 and how these differences translate to greater accessibility for researchers sequencing their libraries through UWBC. The general information about NovaSeq X+ pricing and sequencing depth considerations are also applicable to Parse Evercode Whole Transcriptome scRNA library preparations, which we have recently started offering.
Differences in flow cell capacity/performance for single cell RNA-seq
For a variety of technical reasons, the majority of single cell RNA-seq projects (via the 10X Genomics platform) that come through UWBC have been sequenced on the NovaSeq 6000 using their own flow cell. Depending on the nature and goals of the project, the flow cell options here are:
- SP (2 lanes, 720M paired-end reads*)
- S1 (2 lanes, 1440M paired-end reads*)
- S2 (2 lanes, 3690M paired-end reads*)
- S4 (4 lanes, 9000M paired-end reads*)
*on average
For a read depth target of 70,000 reads per cell (as we recommend for a starting point for most sample types), this works out pretty evenly for pilot studies at the low end for sequencing 1-2 samples of 5,000 or 10,000 cells per sample.
As the number of samples rises, the relatively non-linear jumps in flow cell performance lead to some more tortured experimental design in terms of matching the goals of your experiment with a number of cells per sample in order to plan for a specific read depth/flow cell, and the cost starts to become (more) prohibitive for smaller labs.
At launch, the NovaSeq X+ has one flow cell type, the 10B, which provides an average of ten billion reads from a total of eight lanes. Due to design advances in the consumables for this instrument, the “cost per read” is much lower than the NovaSeq 6000 and it is much easier to facilitate sequencing 10X Genomics libraries on an individual lane or lanes of a flow cell that is shared with other projects at a cost of $2,050 per 1.25B read lane. On the low end, this means (for example) that a pilot experiment of three 10,000 cell libraries sequenced on two lanes of a NovaSeq X+ 10B flow cell will now provide higher reads per cell and cost less than sequencing a single 10,000 cell library on the NovaSeq 6000’s SP flow cell.
On the higher end, if we look at the same 8 sample, 10,000 cells/sample experiment shown for the NovaSeq 6000 above, you can now sequence these libraries deeply for roughly half the price of what an S4 flow cell costs on the 6000.
(N.B.: A major reason why these quotes show that it becomes more cost-effective to purchase your own flow cell above five lanes is because these quotes are for different read length sequencing – individual lanes will mostly be sold at 2x150bp sequencing to be in line with other common library types that the DNA Sequencing Facility handles. The “Full Flow Cell” price quoted here is for the 2x50bp kits, which provide long enough reads for 10X Genomics libraries. Either way, the reads for your libraries will be trimmed to the appropriate length for the application.)
Later this year, Illumina will be releasing the 25B flow cell, which can provide 25 billion reads per run. As the field advances and larger single cell projects start continue becoming more common, this should help bring down the cost of sequencing even further.
Impacts on 10X Genomics Visium Spatial Transcriptomics Projects
As with single cell RNA-seq, Visium projects have generally been sequenced on their own NovaSeq 6000 flow cell, with the read depth being a function of the proportion of the 6.5mm x 6.5mm capture areas covered by the tissue sections. For most of these projects, which have usually been only one or two Visium slides, the SP or S1 flow cells are sufficient (except for the rare combination of a Visium Fresh Frozen project type with very high capture area coverage). In many cases, especially for FFPE (which has a much lower minimum recommended read depth), even the SP flow cell’s 720M reads ends up being excessive. On the other end, relatively large projects with more slides and higher capture area coverage can end up needing one of the higher-capacity flow cells if they want to make sure they have enough sequencing depth for their project goals.
Again the switch to the X+ gives more flexibility and/or reduced prices. Though small projects like the 4 section/30% capture area coverage example will still get more reads than needed, the cost of the NovaSeq run (1 lane of an X+ flow cell versus an entire NovaSeq 6000 SP) will be reduced by more than half. In the middle example, if you’re looking for more than 50k reads per spot for a Fresh Frozen project, one lane of an X+ flow cell gets you there at a savings of more than $4,000 compared to the NovaSeq 6000 S1. Similarly, for large projects, the switch to the X+ offers more granular targeting of specific read depths at a savings of up to ~$10,000.
Additional points for consideration
One additional aspect of the switch to the NovaSeq X+ worth mentioning is the effect on sequencing 10X Genomics Multiome ATAC + Gene Expression libraries. These libraries have substantially different read length requirements and, as a result, on the NovaSeq 6000 the ATAC and Gene Expression libraries must be sequenced separately on their own flow cells. This has meant that even relatively small projects can come with sequencing bills in the range of $20,000 – $40,000. In a welcome change, on the X+ these libraries can now be sequenced on the same flow cell (though in different lanes), substantially reducing the overall cost of these projects.
Lastly, it is worth keeping in mind that for users who have submitted 10X Genomics projects in the past, your experience with turnaround times may be slightly different with the X+. Since purchases of individual lanes mean sharing a flow cell with other projects in the remaining lanes, there may occasionally be increased wait times until a flow cell can be fully utilized.