Should I join the MinION Access Programme?

Oxford Nanopore have recently opened phase 2 of the MinION access programme (MAP) and I have been asked a few times by email by folk wondering whether they should sign up, so here is a little blog post that hopefully will be helpful when making your decision.

In case you are wondering what MAP is: it gives you the chance to try out the Oxford Nanopore MinION USB sequencer for a very small initial outlay, specifically $1000 per “MAP package” (one MinION sequencer and some number of flow cell reagents and sample preparation kits). The $1000 is refundable, although the delivery costs are not.

Joining was a no-brainer for us. We wanted to be one of the first to try the MinION out, we did not care if we spent a lot of time on it, some of which might be wasted, and we had the right resource in place to run it. The MAP also came along at just the right time for us, in the summer holidays. Not having children of school age, summer holidays are the sweet spot for starting an intense project - we are all around, but the students and many of the academic staff are not. Party time.

For other labs it may not be such a clear equation.

The first thing to note is that, in common with other early access programmes, you are helping the company debug its hardware, software and reagents. In other words, if everything worked perfectly, they would be selling you a product, not virtually giving it away.

That’s not to say you won’t generate useful data for your science, but you should certainly not expect it out the box.

What kind of person are you? If you like things to be nice and stable and sorted out, an early access programme may not be for you. In the last six months we have had three chemistry updates (R6, R7 and R7.3), three major library construction protocol updates (SQK-MAP-001 to 003) and countless software updates.

If you like to obsessively practice things until you an expert, this may be a frustration.

What is your level of bioinformatics savvy? Right now there is very little software provided for handling nanopore data, which is characterised by a high error rate (improving, but still not much better than 85% read accuracy on average). The community is grasping its way to solutions for certain tasks (e.g. LAST for alignment to a reference) and SPAdes for scaffolding and repeat resolution along with Illumina data, but right now there is no definitive variant calling pipeline, there is no nanopore-only assembly pipeline and there is no de novo amplicon correction pipeline.

What do you want to do with your data?

If you are familiar with PacBio, you might be tempted to buy into the MAP as a way of getting easier access to a long-read instrument, perhaps for bacterial genome assembly. You can use it really quite successfully to scaffold genomes, currently in conjunction with Illumina data. Right now there is not currently a nanopore-only de novo assembler like HGAP. Hopefully someone builds something soon.

You might be interested in using it for isoform detection from RNA-Seq data, this could be a good usage that plays to the long reads generated by the system that does not rely on single-read accuracy.

If you want a drop-in replacement for a MiSeq or a HiSeq this is definitely not for you. If you want to sequence a whole human genome, think again for now.

So who should join the MAP? right now I think if you want to mess around and be one of the first to touch a completely new sequencing paradigm, to start thinking of experiments that would benefit from long reads and real-time sequencing, then it might be for you. Microbial sequencing would seem to be a sweet spot.

I think it works best for small teams. In our group we have one person (Josh) who can literally do everything, from the sample preparation all the way to the bioinformatics analysis. I would like to think even I could build a library if I needed to, although this might not be desirable for others in the lab due to my general lumbering ineptitude with a pipette. But I can do everything after that point including loading the instrument.

Or to put it another way, if you were to raise the money to buy a PacBio or a HiSeq, you’d probably want some staff to go with it. The barrier to entry for nanopore is significantly (ridiculously!) lower in terms of financial outlay, but you still need to be able to resource producing libraries and running the instrument.

If you would take pleasure in messing around, potentially generating some exciting data, and can handle a bit of frustration, I would say go right ahead and happy MAPping!