Friday, January 05, 2024

2024: A Look Ahead

It's January, and that means the J.P. Morgan Healthcare Conference looms next week -- followed by AGBT just a month later.  Indeed, I've been trying to mark out the "can't miss" talks for AGBT so I can resist over-scheduling them with meet-ups -- but many talks lack titles so that's not easy.  JP Morgan seems to have Illumina, 10X and Nanostring -- and not much else in the way of sequencing-space companies.  But time to prognosticate before all the news happens!


Illumina would probably like to forget 2023, but that's not possible.  Should anyone be foolish enough to remake The Princess Bride, Vizzini should note that one of the classic blunders is to try to close an acquisition prior to gaining regulatory approval.  That led to the unwelcome interest of activist investor Carl Icahn, the ousting of CEO Francis de Souza and certainly nothing positive.

New CEO Jacob Thaysen is a bit of an enigma and an interesting choice.  His previous employer was Agilent, a company with a very different business model. Illumina has focused on genetic analysis via sequencing and microarrays, with most other product lines either serving that or being served by it.  Illumina has traditionally been very innovation driven -- we in the community may grumble that some of that isn't as fast (and certainly much more premium priced!) than we'd like, but I can't imagine anyone wants to trade in a 2024 NovaSeq X for a 2014 HiSeq.  Agilent, in contrast, is a complex conglomerate formed largely on acquisitions.  Certainly Agilent has innovated in some areas (particularly when they can leverage their high density oligo synthesis capabilities), but they don't have one focus.  JP Morgan is probably going to our first chance to see some sort of vision from Thaysen, though maybe not -- this year might be entirely consumed by untangling itself from GraiI.

Some have pointed out that Illumina has a truly commanding position in the field, with hundreds of NovaSeq X in the field and those will probably be workhorses for five years or so.  But if I were Thaysen, I'd be constantly reminding the troops that complacency could be lethal.  The nightmare scenario would be new applications being preferentially grabbed by new entrants.  Illumina's sequencing is likely to be a cash cow for years -- but Wall Street likes cash cows that grow far more than ones that don't.  

Illumina's general problem is how to grow.  Two different major acquisitions have been scotched by regulators on antitrust grounds -- so any attempt to buy a new product line would need to be very carefully vetted and might still prove a risk.  There's also the risk that going into related markets will compete with customers -- who now have real alternatives to flee to.

Element Biosciences

Element's AVITI has a lot going for it -- attractive, open pricing, an energetic and intensely clever team.  So what next?  Element could just continue to focus on supporting AVITI and further tuning the system -- but might AVITI get a bigger sibling sometime that competes in the NovaSeq X space?  

Ultima Genomics

Ultima hosted a few hundred industry folks in the Napa Valley in early December -- plus tours of their factory in the Bay Area -- and I was fortunate enough to be invited.  A full write-up of their technical specs as well as many cool applications that were presented is still in the works, but will come out before AGBT -- where Ultima is Gold Sponsor.  But a few cookie crumbs -- it's 6-8 billion reads per wafer, two wafers per machine and 12 hours per wafer (or shorter, if a counting application can make do with shorter reads).  $100 consumable cost for a 30X human genome.  Variant calling is on par with industry standards -- and if it isn't sufficient there is a "ppmSeq" mode that delivers extremely high accuracy for needle-in-a-haystack challenges.  Plus there where multiple talks showing applications where the question isn't how many libraries do you put on a wafer, but how many wafers do you run per library?  In other words, if you can have 10 or 20 or 200 billion reads cheaply, what can you do that you couldn't do before?  

Singular Genomics

Singular has an interesting box targeted at smaller labs -- the choice of running 1-4 flowcells at a time should be attractive to many core labs with variable demand.  But it's in many ways similar in performance to the AVITI, so they are fighting each other as well as Illumina to grab customers.  The slow delivery of instruments due to supply chain issues in 2022 extending into 2023 certainly didn't help.  Singular went public very early, and is now staring at a significant downside to that: their stock is trading a bit under fifty cents a share -- which is less than the NASDAQ standard of $1.  So far they've only gotten a warning, but at some point they may be forced to run a reverse split (10:1 perhaps?).  Stock splits and reverse splits are a bit of financial hand-waving, but if it must be done it's a distraction. It would hardly be shocking if Singular ended up being acquired or at least did a deal with a big player   

Complete Genomics

Complete Genomics has an interesting product lineup and great technical skills -- but burdened by the linkage to BGI.  Whether it's fear of future trade wars or US sanctions, objections to entanglements in some of the People's Republic of China's less savory genetic projects (ThermoFisher just stopped selling DNA extraction kits to Tibet) or flat out Sinophobia/jingoism, I've encountered more than one lab head who flat out won't consider Complete Genomics.  The rebranding hasn't created much distance -- one AGBT attendee responded to it "Is that what they're calling themselves this year?".  Still, if Complete could publicly bag a significant US or European lab, that might shift the narrative.

PacBio - Short Reads (Onso)

PacBio made a splash last year by acquiring Apton for superresolution technology, but as my conversation with CEO Christian Henry emphasized, that tech won't be in the marketplace in 2024 and maybe not 2025.  So 2024 will likely be a year of just pushing Onso via package deals for long read packages and otherwise no revolutionary updates

PacBio - Long Reads (Revio / Sequel)

On the long read front, will 2024 be the year that PacBio unveils either their benchtop instrument or what I call "the monster" -- a huge instrument for population-scale sequencing that would make the Revio look low-throughput.  And hopefully more Kinnex kits to get higher yields from applications with inherently short inserts. 

There's at least two talks at AGBT on using HiFi data at scale for rare disease diagnosis -- those should be exciting! Could they make more clinicians demand long read genomes in this space?

Oxford Nanopore

Oxford Nanopore is still transitioning users to the Kit 14 chemistry.  Their claims on accuracy keep ramping up and some in the field see those results -- and some don't.  ONT still has an extremely complex product line which they aren't skilled at communicating to novice users -- in Napa Valley I got an earful (made charming by a Yorkshire accent) about the confusing names (that person found the fish-themes about as charming as seafood left in the sun a few days) and trying to match kits to chemistries and such.  But that's ONT's pattern -- great technical vision, mediocre commercial execution.

ONT has been making an interesting push into automation with the TurBOT conventional liquid handler and the TraxION integrated sequencer-and-VolTRAX.  These concepts haven't been enthusiastically received by the research community -- but that's not where they could hit it big.  If these could be "ONT diagnostics lab-in-a-box" enabling diagnostics visions to be easily converted into assays that fit in a pathology lab, perhaps ONT could finally deliver on the promise in that area.  Of course, for TraxION that would mean the VolTRAX technology would finally -- after almost a decade of hype -- be broadly useful.

Ion Torrent

Included here for sake of completeness.  But a non-entity for years in terms of innovation and growth.  They take care of loyal customers, which is the best to say about them

Wild Cards?

Ultima did an amazing job of staying in stealth for a long time; could anyone else.  I've thought about doing another "lazy" round of looking at startups -- there are some out there.  Maybe even some I'm contaminated with an NDA.  Or something innovative will emerge from all the activity in China -- many startups there seem to take the low-risk route of adapting/rebooting older technology, but who knows?  Or with patents expiring maybe the world just needs a solid, inexpensive classic model.  Rothberg's has given some teasers and has some patents out there -- are they building a microfluidic-free, low capital cost compact short read box?  If so, when can we have one? 

Single Cell & Spatial

I keep meaning to track the single cell and spatial fields better, then never quite dig in.  Similar with next-gen proteomics.  10X Genomics is the leader in single cell, but now multiple competitors have instrument-free approaches.  

Spatial seems to be the wild west.  10X has successfully - for now -- knocked out Nanostring at least in certain areas.  But the academics keep coming up with new approaches on a regular basis, so more companies might sprout up.  Maybe Illumina should get involved here -- I just saw a preprint where a NovaSeq flowcell is used to create & catalog spatial indexing oligos, followed by another round of Illumina sequencing to actually read the spatial information.  And certainly more lawsuits will be launched in the complex patent area.

The next-gen proteomics area is interesting.  I've meant to write a piece comparing the proposed acquisition of Olink by ThermoFisher with the just finalized merger of SomaLogic into Standard Biotools (another Grail-like mistake? The lawsuits trying to stop it had just started flying).  I finally got a chance to see an actual QuantumSI device at the Boston Festival of Genomics -- cool technology, but what is the market?  Another incubating draft exists around that.

Sequencing Ecosystem

I already know of one very interesting long read ecosystem instrument that will launch at AGBT.  Long read sample and library prep remains an area with unmet needs.  

As short read sequencing consumable costs keep dropping, library prep will become an annoyingly larger fraction of the total cost.  Does the cost of adapters and enzymes simply have no solution, or is there a way forward to lower costs? Or is the real value in further simplifying / automating library prep so it can be done with less interaction by less highly trained personnel?

Feel Free to Reach Out!

Got a tip?  Want to meet up at AGBT or have coffee when visiting Boston?  Noticed something I omitted or mangled?  Have something interesting you're going to debut at AGBT (and yes, my definition of interesting has a higher bar than most PR agents imagine!)?  It's first.e.last at ("e" my middle initial) -- or find me on LinkedIn or Twitter


Anonymous said...

But do we really care what somebody from the wrong side of the Pennines thinks about ONT software names ?

Anonymous said...

Surely 2024 is the year Roche launches their nanopore machine to clinics?

David Stone said...

Just for completeness on the proteomics side I recommend checking out Nautilus Biotechnology. Very interesting company.

Dale Yuzuki said...

Thank you Keith for all this valuable commentary - nice to see Ultima as a viable alternative to the NovaSeq X.

Interesting that a rebranded Complete Genomics isn't swaying anyone, IIRC it was the Univ Colorado cancer center (Anschutz) that recently installed a T20x2 monster, with $0.99/Gb pricing (and a price of about 1/2 a NovaSeq X to boot). I've heard there are a few T7's in the US as well; how many would need to be installed (in your opinion) to sway that bias against anything from China?

In a world where just about everything can be traced back to some part from China, I think it's a bit hard to exclude MGI / Complete Genomics out of the competitive landscape. 2024 will be shape up to be an interesting year for sure. Thanks again.

Dale Yuzuki said...

Hi Keith, thank you for the valuable commentary! (Sorry I'm late to the party...)

Surprised (and pleased) to see that Ultima is going to give the NovaSeq and Illumina some real competition, having over 10 customers already (and AGBT around the corner) sounds like they will be "the" company to watch in 2024.

On the Complete Genomics front, how many systems would need to be installed (and what proof statements would you need additionally) to take them as a 'serious' competitor? I understand they sold a T20x2 to Univ Colorado, and have a few additional T7's in the US. I certainly understand the concern about the 'blacklist' as well as the sleight-of-hand in 'manufacturing' these CGI systems in NoCal, but at 1/2 the price of a NovaSeq X and 1/2 the consumables ($0.99/Gb pricing), it would seem too good to be true...

Anonymous said...

You wish, vapour ware is this mishmash of genia and Stratos.... Perfection is the enemy of done... In the case of this sequencer. The post is quite telling that they're not even a footnote on this post