Even before new CEO Francis Van Parys took the reins at Oxford Nanopore with the start of the month, the company had made yet another priming to the instrument lineup. The sample-to-answer ElysION has been sent to its eternal rest, presumably in the nicer part of Hades
Image courtesy of Rasmus Kirkegaard
ElysION started as project TurBOT, and was intended to enter applied markets which required simple end-to-end operation. These were envisioned as existing in biopharma QC, particularly for RNA therapeutics, food safety, biosurveillance, and so forth.
ElysION was always the antithesis of longtime CTO Clive Brown’s instrument aesthetic. Brown always spoke of wanting “no kit”, a vision of any individual generating data in the field with no specialized training and no laboratory equipment. ElysION was a mid-to-large liquid handling robot that wasn’t ever going to be run in an jungle camp.
ONT contracted Tecan to build the instrument - which meant the library construction support was an easy job since Tecan has extensive price in the NGS library robot space. Loading the flowcell was the tricky part - getting precise alignment to the loading port is not super hard, but there is the matter of manipulating the gate that protects the inlets. Stuff and sometimes balky with human fingers, it was certainly never intended for robotic manipulation. Tecan solved this with a clever jig tool that the robot can pick up and open and close the gate, but so much complexity to cope with an early ONT platform design decision.
But once built, ONT failed to rapidly deploy applications for ElysION. If you are going to build a turnkey system not aiming at hackers - and actively disdained by many of them - you’d better crank out keys to turn!
The lack of focus on the commercial side was clear at last year’s London Calling, where I stood looking at the display model for At least 10 minutes whilst a trio of ONT sales folk nattered away amongst themselves. And once I could get somebody to explain the status, it seemed it was still only supporting the rapid kit that had nearly no steps and Brown pitched could teach any C-suite driving to run.
A telling sign was when an academic group excitedly tweeted out last fall that they were the proud owners of the twelveth ElysION - this after a bit over a year since launch. If the goal of ElysION was to sell flowcells and kits, that wasn’t going to move much consumables at all.
ONT's Growing Graveyard
Graphic courtesy of Rasmus Kirkegaard
ElysION will have plenty of company after tipping Charon. The dedicated, confused laptop for MinION is gone. A grassroots campaign to save the P2 Solo got only its support extended. VolTRAX leaked its last oil quite a while ago. You can no longer finagle a Flongle. And that’s just the released or nearly so stuff (and I'm sure I've forgotten some).
Save the P2 Solo buttons brought to AGBT by Graham Wiley - I'm responsible for distributing a trio
There’s Plongle / ONT’s other idea for docking with the high throughput automation world with a 96 well sequencer. The next generation mini flowcell SmidgION which would support a thumb drive sized USB adapter/sequencer. Ubiq to allow anyone to run their own cheek cells. TraxION - the other integrated sample-to-answer concept. Solid state Nanopore is another once-touted research initiative that seems to have been iced.
On the chemistry side, there was the Cas9 Kit for capture sequencing - reputed to be flaky but on the other hand rival PacBio seems to be doing well with their Cas9- based PureTarget kit. Also the lyophilized field kit for rapid preps.
ONT was an idea factory with many exciting ones - many more than their commercial side was able to properly promote and support, but now the focus is on getting to break even and cuts must be made. Perhaps some of these might later see light of day - solid state nanopores - but most will likely never be heard of again.
Will there be more pruning with a new sheriff in town with no prior emotional investment in the company? ONT is now marketing two partially incompatible product lines. There’s MinION and GridION on the side, and the three PromethION devices (P2i, P24, and P48) on the other - it would certainly be simpler to squash down to a single line. GridION and P2i represent a bit of product overlap for many potential customers - and overlap can breed customer confusion. P48 reputedly has difficulty keeping up with basecalling if really busy - and I’ve heard this doesn’t require having all 48 flowcells going
Nava Whiteford has previously outlined an economic case for focusing entirely on the PromethION line. That would be a radical shift, nearly completely abandoning the idea of mobile sequencing
ElysION Redux?
ONT could of course do nothing; ElysION and all its technology confined to an oubliette. Simple, efficient - but no attempt to recover some value from their investment.
Or they could either try to license the comments or or just open source them. This might be complicated by whatever contrarian terms exist with Tecan. But I certainly know of one party interested in the idea (same guy, coincidentally, regularly walks my dog).
Or, ONT - or perhaps someone else - could try a different solution. Ideally future flowcell casing designs would be automation-friendly, but perhaps in the near term a simple hack of manually opening the port and then adding an automation-friendly lid. No idea if that is really feasible, but someone mechanically clever might come up with it.
I would generally think that many future integrated sequencer designs will look more like a bolt-on module to go on a large deck liquid handler or a very minimalist unit to attach to an integrated workcell or autonomous laboratory. ElysION’s huge footprint and minimal utility certainly wasn’t a winning combination. That could also be a consideration for future instrument designs - it would be painful to burn a bunch of deck space for a PromethION device when all that really must be occupying space is the flowcell and connections.
Parthian Thoughts
ONT continues to be a company in transition and increasingly trying to be mature and lean. London Calling is less than two months away, and should be very interesting - particularly as the debut of Van Parys. Hope to see you there!
No comments:
Post a Comment