London Calling 2025: What I'm Thinking About

London Calling fires up for real on Wednesday; Tuesday has the training courses I haven't signed up for.  I don't have any grand predictions, but rather some thoughts of things I am trying to keep my antennae particularly tuned for.  Sly tips of course are always welcome I can be DMed on Discord, LinkedIn, X, or email me at keith.e.robison at Gmail.com..

Before I get to that, there is one observation about the meeting itself that fits with the turbulent year ONT has had, with Clive Brown leaving and CEO Gordon Sanghera openly talking about coming to grips with the reality that they might be taken over at some point.  Last year ONT tried having a third day of talks, devoted to industrial applications, with attendance requiring an extra charge.  It was much more sparsely attended, but still a healthy crowd.  This year they again announced that such a day would occur, and then when the registration opened for the main meeting there was a promise that the industrial day would have details available at a later date.  But that date never came, and eventually (long after I had assumed it dead and planned to fly to ESHG in Milan that day) the extra day disappeared from the announcement - though with typical ONT lack of attention to details there's still the fossil of it in "We are delighted to be welcoming you back from 20 - 23 May 2025" - since the 23rd is the day you could never register for.

How Will Clive Be Officially Noted?

I wrote a long piece yesterday on Clive Brown - which almost immediately gathered some deliciously spicy comments (plus a correction - I misremember the original format as POD5 when it was FAST5) - with some heat pitched ONT's way and some directing derision at my grasp of the history.   How will ONT note his passing?  I doubt (and would be quite bothered) they will unperson him ala The Commissar Vanishes.  But will he receive some sort of celebration?  Will he be in attendance and invited on stage as an emeritus persona?

Will We Hear More About Acquisition Vulnerability?

On the one hand it was surprising that CEO Gordon Sanghera even publicly discussed the possibility of being bought out; ONT has been fiercely independent in culture to the point of being dangerously insular.  Will Gordon mention these comments in his opening remarks?  

Will The New Nanopore Competitors Be Mentioned?

ONT is no longer the only nanopore platform being talked about.  There have been some obvious knockoffs in China, with even the physical forms looking close to ONT's.  But now there are potentially legit competitors - though probably ONT will say they've swiped IP.  Much attention, such as from this space, has been poured on Roche's SBX announcements before and at AGBT, and then there's also BGI's CycloneSeq which is much closer to ONT in concept. 

For one thing, it creates the semantic question of how to refer to these platforms - I'm certainly in the lazy habit of calling ONT sequencing "nanopore sequencing" as a synonym.  So now perhaps we should all be specific to ONT nanopore or BGI nanopore sequencing, except when comments are truly applicable to both.

SBX is clearly a different beast - it uses nanopores for signal acquisition but that signal arises from library molecules which have been resynthesized with the funky SBX nucleotides.  That's an obvious bashing point for ONT: SBX as currently sketched out won't have native modification detection but must rely on chemical or enzymatic conversion methods. Though Roche has hinted they might have a way to get 5-methylcytosine in the same reaction.  

ONT for a long time tried to avoid referring at all to their competitors, and then switched to very oblique references.  There's always a danger that even a pointed critique of a rival directs the spotlight to them - let alone possible litigious complications.  So how will ONT describe the rapidly changing landscape of sequencing technologies? 

Product Line Streamlining?

ONT's product line has sooooo many options, many incompatible with others.  Buy the wrong combination of library prep and flowcell and you are not moving your experiment forward. Maybe that's now down to Direct RNA and Flongle flowcells.   It's been years since I was involved in consumable purchases, but we also had problems in 2020 +/1 a year with requesting one chemistry and having a different one shipped.  

The problem is that every one of these variations in instruments, flowcells and kits has a devoted following, so discontinuing any is sure to raise complaints.  Only a few "dead is dead" items appear on ONT's Discontinued Products list.  There's the Cas9 enrichment kit was axed over a year ago (and sources tell me never worked very well).  There's also the PCR Sequencing Kits, which went away almost two years ago.  If there was some justifiable complexity, having a good way to convert PCR amplicons easily to full length nanopore reads would seem  a clear case.

So what will be the net effect on product line complexity this year?  Some culling?  Stasis?  Or further complexity?

Also on the complexity front: the new flowcell scheme with the little dongle devices and a flowcell port designes for easy pipetting by human or machine.  Is that rolling out - it has clear possibilities.  But it also seem to be taking over the Flongle's niche, so would it drive Flongle to extinction?  

ElysION

ElysION is ONT's Tecan-engineered benchtop robot for library preps, intended to serve markets that want to run a few workflows routinely.  In theory, it's just set the deck up and go.  ElysION even loads the flowcell itself, using a set of complex movements and specialized jigs to fight the same MinION rotating cover that has bedeviled many a human fingertip.  

ElysION last year sported a pair of MinIONs, but the intent is to be able to mount a P2 within the robot to provide higher data volumes for applications which require it.  

I'll be looking to see which applications ONT highlights and which they have end-to-end protocols - which especially includes informatics since they want to sell these into environments that are the polar opposite of biohackers - industrial firms that want to train the operator minimally and also be able to have that operator walk away from the machine after the run is launched.

Who's Giving the Big Talks?

Long ago in the days of yore, every meeting featured a marathon talk by Clive Brown going for an hour or more, followed by Q&A (and annoyingly, the Q&A never makes it in the recordings).  Clive would be exhausted and hoarse at the end, but all sorts of nuggets were dropped to thrill the faithful (and give hacks like me material to write about).  The first modification was to have Rosemary Dokos police the product release announcements after any of Clive's excessive promises, and then later it would be a set of high lieutenants who would cover different segments, with Clive usually slipping in for a bit.  That also provided a seamless succession path.

So who will be giving the big product updates this year?  Will we be seeing any new faces?  Have any of the prior faces left - one issue ONT has been coping with is that many of their talented R&D folks have struck out with their own startups.  Interestingly, of the ones I can think of none of those startups directly mesh with ONT technology.

Will There Be Any Blue Sky 

Of course, the most exciting bit would be Clive's talk, where he would drop some crazy idea.  So trying to sequence Clive onstage with Ubiqibopsy or Ubiq for short (note to branding: check if there is a well-known dystopian novel by that title).  Sequencing within the VolTRAX. DNA data storage with MinION - which I hated, still hate & take perverse pleasure in the fact that it is clear my screed against it got under Clive's skin as judged from tweets where he has complained about "some people hating it".  There were also the very long range, game changing new technologies - voltage sensing or solid state pores.

So what will we see of the high concept, blue sky ideas?  Are any of them actually near release, or are they doomed to always be (like monorails) the technology of the future.  Will any of the old edgy stuff be mentioned again?  Such as sequencing chromosomes end-to-end as a single fragment, or cell free DNA liquid biopsies (which at one point ONT was promising shareholders was the road paved with golden bricks), or the Plongle 96-well sequencer?

Any Real Details on Cepheid Collaboration?

The biggest commercial news of late for ONT is their collaboration with Danaher-owned Cepheid on microbial diagnostics.  So will we see any real details?  Heck, I'd be happy to have a slide of "everything Keith got wrong".  Any mockups or concept art of how ONT will be incorporated or perhaps suggestions of which assays might be ported?

Any suggestion of progress on other collaborations - or any new ones - would also be interesting.  Historically ONT was not one to partner, but that seems to be changing - but will the trend continue?

What Is The Path To Breakeven?

ONT is now a public company in late adolescence; it can no longer subsist on buzzy technical announcements alone.  They and their customers have pioneered so many fascinating applications, but few have risen above tiny niche status.  Plasmidsaurus (I'm a happy customer) is one obvious example, and CaribGenetics would appear to be another.  There are probably a handful more - companies that are selling ONT-based services to customers at scale.  But will they be more?

For example, since the first community meeting we've heard of compelling tales of using nanopore sequencing for rapid diagnosis in a way that should have significant gains. At that first New York meeting, there was the example of rapidly diagnosing severe aneuploidies in embryos from in vitro fertilization. There, fast sequencing avoided freezing embryos - and also meant the couple could accomplish the whole process in a single visit.  Other times its been rapid diagnosis of leukemia fusion genes to enable smarter therapeutic selection.  Again, a key benefit is enabling diagnosis and prescription to be in a single treatment, a serious advantage in the face of many patients in the US living far from specialized medical centers - a trip to such a doc means lost wages, possibly childcare arrangements, travel costs and lodging.  Or the really exciting pilots from Matt Loose and collaborators of demonstrating the possibility of making calls from nanopore data that will alter how a brain surgeon tackles a tumor - and doing so while the patient is being worked on by the brain surgeon!

But that's also perhaps the snag in all this - in an ideal world those out-of-pocket costs would be fully recognized as a target to be reduced.  But the payor for the medical procedure - a public health system (most of world) or private insurance (US and some other nations) - isn't typically paying those and so isn't strongly incentivized to reduce them.  

But in any case, no change in the system will occur without a great deal of grunt work - getting payors on-board, getting regulators on-board, showing therapeutic advantages in sufficiently large numbers.  And those slogs are the kind of thing that ONT has shown strong interest in committing too - they are just not buzzy enough.

There's also entrenched interests to battle, interests who will win by default if not fought.  Any decentralized diagnostics platform can be seen as a threat by established centralized facilities.  There's also just general skepticism of whether these approaches really move the needle, or if they are exciting but in the end no better than the boring and existing.  Again, fighting these requires carefully planned campaigns, not one presentation followed by "ooh, a new shiny toy"

Short Reads

ONT has sometimes tried to play in the short read space, but without much enthusiasm.  Early editions of the platform behaved poorly with short fragments, but changes were made to enable these.  But are there any compelling use cases?  One would be direct RNA sequencing of tRNA and various regulatory and structural small RNAs; tRNAs have a wealth of base modifications and who knowns how much modification is on all those other small RNAs. 

But short DNAs?  Perhaps useful in some low resource or field environments?  That's the problem - there's a dearth of compelling stories of how nanopore might bring advantages here, since the cost per fragment and number of fragments generated will likely be far worse than even a low end short read sequencer.

The T2T Bundle

Last year there was huge hoopla around a bundled set of kits to sequence mammalian genomes (or similar sized ones) telomere-to-telomere (T2T).  The scheme required three libraries - a standard library which would be the workhorse, an ultralong library and a Pore-C library to establish connectivity through extremely long repetitive regions.

At the time, the excitement from the Nanopore executive team seemed to be balanced by community skepticism.  The cost of data generation was many times that of assembling a very good but non-T2T genome with either just a standard ONT library or with competing technologies.  So was it worth a substantial premium to get T2T?  Or put another way, would you rather have 1 T2T genome or perhaps 3 not-quite-T2T genomes.

The gap has been narrowed further - a preprint from the HERRO group claims that with the newest assembly software human genomes with many T2T chromosomes can be assembled from a standard ONT library.  That's of course good for ONT, but it would seem to further doom the T2T bundle - which I can't find on the Nanopore Store.  If I search for it, I find the original press release.  Will ONT try such marketing bundles in the future, or is the field just too fast moving for such to work?

VolTRAX Dead or Mostly Dead?

The VolTRAX electrowetting technology was always a favorite of Clive Brown's, but as other companies have found is a difficult tech to master. VolTRAX as coupled to MinION sequencing also had the problem of not a very compelling value proposition.  For example, one kit ran the transposase-based rapid prep on a handful of samples, something that isn't terribly challenging at the bench (says the person with 1 usable nanopore read to his name). VolTRAX also existed in a grey zone - still too complicated for a rank amateur - especially since you still had to perform the tricky task of loading the flow cell - but not streamlining enough or working at interesting scale for the expert.  VolTRAX was great at demos spelling its own name, but rarely rose above that.

Last year ONT pulled the commercial listings for the device and kits.  So is VolTRAX truly dead?  Or is it only mostly dead, waiting for its own Miracle Max.  Clive had proposed TraxION, a VolTRAX truly married to a flowcell, though how the prepped library was to be transferred seemed to still be in "and then a miracle occurs" level of details

Protein Sequencing

ONT has made noises about developing nanopores for sequencing proteins, building on the work of Jeff Nivala which was presented a few years ago.  It's an exciting idea in a hot area, so it will be interesting to see what if anything is presented this year on progress here.  I have a bunch of other thoughts on the subject, but I'm going to save those for a long piece on an idea I've kicked around for years but now am going to force into the open - probably tomorrow just as the conference begins.

What have I missed?  What do you think I should be looking for? What have I written above that's utter rot (there's one of those comments on the Clive Brown piece too)?  Please let me know in the comments!