Illumina Would Like to Change the Conversation

A maxim from the great but fictional advertising executive Don Draper: "if you don't like what people are saying, change the conversation".  In an online strategy update presented two weeks ago ( Slides / Replay  ), Illumina announced they'd like a new conversation around sequencing costs.  No longer will they tout reagent cost per basepair, but instead will be focused on the total cost of sequencing workflows.  The obvious cynical response is that Illumina is conceding defeat on the raw cost, having been severely beaten by Ultima Genomics (and Complete Genomics aka MGI, but that group continues to face stiff headwinds) and even matched - if you have the volume - by Element Biosciences.  Total cost of ownership is what really matters, right?  The catch is how is it being calculated and who is doing the calculating?

It has always been known that cost per gigabase or per million reads was a convenient fiction.  Convenient because only simply arithmetic was required to convert performance specs and list prices into the metrics.  But a fiction since all the other costs didn't magically go away.  But which costs are we now counting? And how do you count them?  For example, if the library prep requires 4 hours of hands-on time, whose hands?  A Ph.D. paid at Boston rates or a fresh B.S. graduate paid at U.S. heartland rates? (not knocking either - but cost-of-living in Boston is particularly painful for those starting out and that is reflected in higher wages). Illumina would particularly like to highlight the value of their DRAGEN computational acceleration platform - but when comparing it to conventional compute, what number do you pencil in?  It all runs afoul of a dictum thrown out at a class on product financial modeling back at Millennium: keep it simple - "why spend the effort to invent a lot of numbers when you can just invent a few?"

Illumina would like to calculate from having a purified DNA sample to results on the other end, which fits with their strategy of offering - but not insisting on - vertical integration.  So library prep, running the sequencer, primary bioinformatics and secondary bioinformatics.  The same webinar teased that two new library prep products will be coming, though a year to a year-and-a-half (if they keep schedule) away that will further fit this model.

Other companies have already been taking potshots at Illumina on cost angles that might not make it in Illumina's official numbers.  For example, Ultima Genomics UG100 has a "daily care and feeding" arrangement which differs greatly from Illumina's "load a new run after the next has finished" - since Illumina runs often annoyingly exceed an even multiple of 24 hours, full Illumina instrument utilization will ultimately require night and graveyard shifts.  Oxford Nanopore would similarly tout the ability of PromethION to launch new runs at will.  Element and Oxford would both count to lower capital costs.  And so on.

Which also brings up under what scenario are we calculating costs?  One with enough samples arriving all-at-once to get maximum cost efficiency on a NovaSeq X 25B flowcell?  Or a scenario favoring Element where you must run now with a much smaller batch of samples - which seems to be a more practical model for the majority of core labs.  So many ways for each company to frame the problem to favor themselves and prevent any sort of apples-to-apples comparison!

Two New Library Preps -- in the Future

Illumina touted two new library prep approaches they are developing - one which claims it will perform library prep on the flowcell and another offering "5 base" sequencing which would call 5-methylcytosine (5mC).  No details were provided as to how either of these would accomplish this.

Element has been leading in moving library processes onto the flowcell, though in their case it isn't the initial library prep but hybrid capture enrichment.  The Illumina prep won't be cost feasible without some sort of pre-instrument operation; the input DNA's must be tagged because there are just about no applications which call for running an entire 25B flowcell on a single sample.  Perhaps this would just be tagging with barcoded Nextera (Tn5), but then the samples can be pooled and placed on the flowcell to complete the process.  Another speculation I've seen is that the PIPseq templating technology acquired from Fluent would somehow apply.  

Illumina not only is promising a simplified workflow, but also that the quality of the final data would be better than any other solution out there - and they were clearly aiming at (but without naming) PacBio HiFi data.  That is certainly in the category of "show me the data!", as that is a very hard challenge - particularly since good long range contiguity data requires high molecular weight preps going into the process.  This claim might suggest they are using the PIPseq technology to generate linked reads ala the old 10X Genomics kit - but I still remain skeptical that such data can deliver in the face of certain types of repetitive content, such as Variable Number of Tandem Repeat (VNTR) alleles where the repeat array is longer than the actual read length.  And there are a range of applications - perhaps not yet as big as whole human genomes but someday - which require high accuracy single molecules - each single molecule read is the datapoint.

The other big promise is a 5-base reading chemistry.  The first thing to note is it isn't the same as the "on instrument library prep".  Illumina also didn't talk about reading 5-hydroxymethylcytosine (5hmC), the rarer but potentially buzzier additional mammalian epigenetic mark.  The claim is their method will be a simple workflow with a single library, so not a case of running one bisulfite or enzymatically modified library to read 5mC and another native one to read the genome itself.  A speculation I'll throw out is again around PIPseq - perhaps some partitions would have the enzymes to recode 5mC to something else (or all the non-5mC to U, as most modification methods do.

The most advanced approach in this space is Biomodal, which is overdue for a focused approach (and was founded by the creator of Solexa technology, Shankar Balasubramanian, originally under the name Cambridge Epigenetix).  Biomodal creates libraries which effectively are duplexes, with one read reading one strand and the other reading the other.  By clever series of enzymatic steps, the end result is that comparing the two strands can reveal both 5mC and 5hmC while still reading the underlying sequence - 6 base sequencing.  Of course, there ain't no such thing as a free lunch - any advantages of having paired end reads for mapping are no longer available, and there's always the danger of creating noise by the enzymes not always hitting their marks. 

Illumina didn't announce a purchase of Biomodal, so they must have found a different way of converting.  They also promised a simple workflow - a knock I've heard on Biomodal is the workflow is not simple.  

One smaller tease from Illumina is a goal of putting XLEAP chemistry on the MiSeq - which would certainly tidy up their product line.  But would this be existing MiSeqs or is a next generation MiSeq under development?  That was left ambiguous - as well as what would happen to MiniSeq and iSeq in the process.

All-in-all, it is a welcome change to see Illumina acting as if competition exists - the webinar was full of claims that the company is listening to their customers and seeking input.  So they are going to talk the talk of not being stuck in monopolist mode - but will the walk the walk?  Let's see how the next few  years play out

Musings on Possible Fixes To PacBio & ONT's Achilles Heels

I recently tried to place a claim that I had first conceived Oxford Nanopore's "6b4" strategy for solving homopolymers, but that appropriately brought a number of citations for the concept that predated my blog piece.  Not one to give up easily (and as hinted in that piece), I'm going to spend part of this piece trying to stake claim on some new concepts for fixing Oxford Nanopore's homopolymer issues - and PacBio's trouble with polypurine stretches.  To be honest, much of this piece will consist of me posing questions I haven't bothered to try to chase down if they've already been answered in the literature.  But not only might someone do that, but it may well be that data already exists in the public sphere to explore proof-of-concept!  But I haven't checked that either - though doing so was on my list of "what to do if management gave me the summer off" - but they didn't.

Tagify: seqWell's Line of Tagmentation Reagents Awaits Your Creative Thoughts!

One of the most important enzymes in the sequencing world, one which enables spectacular creativity on the part of novel assay designers, is Tn5 transposase.  Personally, I spend many times each month thinking about how to use Tn5 and its ability to tagment - both tag and fragment - input DNA. There’s even reports that Tn5 can tagment RNA-DNA hybrids such as from reverse transcription or even long single-stranded DNA.  I’ve covered seqWell in the past,with their fully kitted reagents; now the company (which just turned ten) is launching a Tagify product line that is focused on enabling NGS dreamers to easily explore new Tn5-based library preparation methods.

mRNA Therapeutic / Vaccine Quality Control: A Major ONT Opportunity?

Oxford Nanopore is in the process of morphing into a product-focused company, and so must identify specific markets in which they believe nanopore sequencing can compete or even dominate.  One such market that was spotlighted this year at London Calling is the quality control of mRNA therapeutics, where nanopore sequencing may be able to replace a kitchen sink of technologies and often provide superior data.

Pharmaceutical and diagnostic quality control is both similar and very different to research.  While many sequencing research experiments are to some degree a fishing expedition, in a quality control assay very specific hypotheses are tested with specific, pre-determined thresholds.  Consistency of results is the most critical; an assay run today must be comparable with one run last month or last year.  These markets may be less sensitive than research to cost; if a QC test is part of qualifying a vaccine batch which will sell for millions of dollars, spending a thousand on that assay isn't unreasonable at all.

It's worth reviewing the process of how mRNA vaccine drug substance are made. The initial vaccine design is synthesized into a plasmid; this design includes a poly-A tail followed by a restriction site (which cannot occur within the vaccine design, though it could occur elsewhere in the promoter backbone).  Enormous batches of plasmid are grown in E.coli and extracted and then linearized with the restriction enzyme that cuts after the poly-A tail and has no sites .  In vitro transcription is used to transcribe the linear template, with the nucleotide mix containing a uridine analog such as 5-pseudouridine in place of uridine.  If the BioNTech process is used, then the nucleotide pool also contains a guanine analog which contains a 5' cap structure (CleanCap).  If Moderna's process, then the in vitro transcription product is treated with a capping enzyme (typically Vaccinia Capping Enzyme aka VCE; please see conflict-of-interest disclosure at the bottom of this piece). After purification and concentration of the active drug substance (removing nucleotides, process enzymes, uncapped product, etc), drug product is ready for the finish-and-fill steps of encapsulating it in the lipid nanoparticles and filling vials for distribution.

QC is all about detecting what might go wrong and ensuring consistency of product.  mRNA therapeutics and vaccines are complex products, with many possible parameters to measure.

First, there's the question of "is this the right product?".  mRNA vaccines continue to evolve and expand in scope, with new designs targeting specific SARS-CoV-2 variants, influenza and RSV vaccines.  If a vaccine product should be one specific variant, it is mislabeled and unusable if it is really a different variant.  Many vaccines are now polyvalent, targeting multiple viruses or multiple variants within a single virus.  This adds a whole new dimension of not only have the correct set of vaccines been blended together, but is the fraction of the whole for each one within defined bounds.  As RNA products, there is also the question of whether the RNA is what was intended and no mutations have arisen during propagation of the plasmid.

Similarly, was the correct uridine analog used in production?  In vitro transcription may generate undesirable products, such as double-stranded forms of the intended product.  How much of these are present?  What fraction of the transcripts are capped? Are the RNAs full length or are there partial or degraded versions present? How much plasmid is left, and is it linear or closed-circular form?  How much E.coli genomic DNA contamination is present?

Many "old school" technologies exist for many of these questions.  A standard gel can be used to assess the length distribution.  Sanger or short read sequencing can be used for sequence verification - though Sanger will be a poor choice for multivalent designs.  HPLC may be used for a number of the questions.  But typically each assay asks a single question, and often with significant constraints. For example, if a problem is discovered in a multivalent vaccine in which there are out-of-spec shorter RNAs present, can Sanger or short reads tell which component is degraded?  

Pfizer has published an approach using specific RNA cleavage (harking back to how Woese sequenced RNA to create the Archea hypothesis - and much before) feeding into mass spectrometry.  In some ways it looks like really short short read sequencing - some fragments are indistinguishable.  The perceived advantages are that this method can distinguish fragments with the correct uridine analog vs. those with just uridine and it can distinguish capped 5' end fragments from uncapped ones.  I've meant to do a deep dive on this for over a year after Kevin McKernan had pointed me to it; time to re-prioritize that!

ONT is proposing that Direct RNA sequencing (plus DNA sequencing of plasmid batches) can be used to build a single assay to test nearly all - if not all - of the final drug product and standard DNA sequencing for assessing batches of circular or linearized plasmids.  As noted in my piece on ElysION and TraxION, this sort of "applied market" would be very appealing to ONT in terms of providing a steady source of revenue.  Direct RNA is the only currently marketed sequencing approach that can look at the modified bases, potentially giving ONT a large edge.  Many of the questions of interest are better answered with long reads - the distribution of RNA species lengths, which RNA species are which length - giving any long read platform an edge.  Should there be a problem, long read sequencing can quickly identify correlations between different anomalies.  

Of course, this does require levels of precision and accuracy.  Data was presented suggesting that minor variants can be detected at around 1% frequency.  Improved algorithms for poly-A length determination appear to enable very precise determination.

ONT dreams of covering more angles.  For example, nanopore sequencing on its own probably can't determine whether the 5' cap structure is present. But, with some sort of pre-processing - perhaps resembling Cappable-Seq/Recappable-Seq, it may be possible to tag either correctly capped or non-capped messages.  Similarly, it may be possible to differentially tag single stranded and double-stranded RNA

In terms of scale, Direct RNA sequencing in the current ONT protocol cannot be barcoded.  For huge infectious disease batches that may not be an issue; for small personalized cancer vaccine batches cost may be more of an issue.  Flowcell washing may be one solution, or ONT may be driven to enable barcoding (there are apparently external protocols for this).

How big a market will RNA vaccines be for ONT?  That is of course the big question.  mRNA vaccines seem to be here to stay, but how many more vaccines will be launched?  Delivering other therapeutics by mRNA is still an unproven market.  If mRNA delivery turns out to be a growth market, ONT can ride that wave.  If it remains a niche market, there's still gain for ONT but not what will drive them to profitability.  Lacking a reliable crystal ball, everyone must simply wait to see how this unfolds.

Conflict of Interest Disclosure / humble brag / me pretending to do Business Development.  I am (still!) employed by and hold stock in Ginkgo Bioworks. During the pandemic Ginkgo Bioworks developed a new fermentation process for producing Vaccinia Capping Enzyme (VCE).  This process is ten-fold more productive than the baseline process.  Ginkgo licensed this process to Aldevron, which is now owned by Danaher.  So production of mRNA therapeutics with capping using VCE may, through an opaque process, benefit me financially.  Little to no evidence of that so far, but it could happen!  And if you have a fermentation process that could be tuned up, feel free to reach out to me!

ONT T2T Genome Bundle: Hot New Thing or Flash in Pan?

Last month at London Calling, Oxford Nanopore announced a consumables and reagent bundle which enables generating six telomere-to-telomere (T2T) human genomes for $4K each.  Even in the very friendly audience at London Calling, there was some skepticism over the market viability of this offer - how much would it really drive sales?  T2T human genomes really only became possible in this decade.  The first examples of T2T chromosomes generally used a mix of different technologies, often including PacBio HiFi, ONT Ultralong and BioNano Genomics mapping information.  What ONT is proposing is the ability to routinely generate T2T genomes using only ONT data.  

ScalePlex: Easing High Sample Count 3’ scRNA Sequencing

Scale Bioscience officially rolled out today - their rep was already talking about it at the Boston Single Cell Symposium I attended yesterday - a new cell indexing reagent called ScalePlex to streamline single cell 3' RNA sequencing of multiple samples.  

Aftermath

I have multiple drafts of posts trying to finish up my London Calling items and then a long list of ideas in various stages of gestation - and been dangled a new tech update under embargo.  But today, I'm on a mission - to help my now former colleagues.  My employer, Ginkgo Bioworks, has executed an approximately 25% layoff.  I survived the cut, but the list of talented, wonderful people who have been cast away is long and covers a wide range of talents.  You really could start multiple quality small biotechs with these new unemployed people.

I've been laid off twice before and it's miserable.  I was lucky each time and had only a short period of unemployment - but biotech was doing well each of those times.  The industry is in a serious slump right now, with many companies cutting back and some closing altogether.  Even large companies are slashing away - Takeda is setting free over 600 employees here in Boston - perhaps some are remnants of the Millennium acquisition.  Far too few companies are being created.  

So if you have leads on open positions, I am listening.  You can leave comments, email me (keith.e.robison on Gmail), connect on LinkedIn, DM on Twitter, etc.  I've never received a message by carrier pigeon, but if that's you're style I won't object.  All will be passed on to a Ginkgo alumni community.

As a meme I saw put it "this too shall pass - perhaps pass like a kidney stone, but it shall pass".  The long-term societal upside from biotechnology is too great for this to be anything but a temporary dip - but temporary can be a very long time.

CariGenetics: Breakthrough Breast Cancer Genetics in the Caribbean - but Also a Template for ONT Clinical Push?

London Calling isn't nearly as exhausting as AGBT, but the first day of talks is packed and then follows with the social event that goes late - this year with CEO Gordon Sanghera living out his dream of being the frontman for a band.  Then if you'd like you can follow the crowd to a pub to drink on ONT's tab (that and crashing the ONT wrap-up dinner is the extent of my drawing personal benefit from ONT, contrary to a commenter on the prior piece who wrongfully believes they fund my LC expenses), and when that pub closes to another one (I peeled off after the first pub).  So one can be a bit draggy heading into the second morning, but that was solved quickly by CariGenetics CEO Dr. Carika Weldon, who wowed with an exuberant strut down the central runway to a lively calypso beat - and then wowed everyone further with a stellar presentation.  She also gave the lunchtime Product Demo talk (alas, I can't find that talk either on YouTube or in Nanopore Community) in the central product area, filling in some colorful details on her young company's early travails - all resolutely conquered.

ElysION vs. TraxION: Divergent Shots at Applied Market End-To-End Automation

London Calling was a particularly good opportunity to take stock of Oxford Nanopore's progress to a "fire-and-forget" sample-to-answer solution for "applied markets" such as food safety, public health and biotherapeutics quality control.   ElysION (formerly Project TurBOT) and TraxION represent very different approaches targeting different subsets of this broad market opportunity - and I heard from some interested parties that neither is quite what they want.  That doesn't mean they aren't right, but it does mean ONT may need to think of more approaches.

The broad concept is to have a a device that takes some sort of biological input, with minimal to no upstream processing, and performs all necessary steps so that nanopore sequencing data emerges from the instrument, with no human intervention after the run is set up.  ONT is envisioning these being placed in clinical labs, public health labs, biotherapeutic quality control labs, etc.  

Thoughts on A Decade of Oxford Nanopore Sequencing

I'm writing this the eve of Oxford Nanopore's London Calling conference.  This is a big one, as this summer marks the 10th anniversary of ONT releasing devices into the wild.  It's been a long, interesting journey and I'm much too jet-lagged to try to review old posts or even link to them, but a bunch of thoughts have been in my head the last few days. 

FOMO Index at All Time High This Week

I'm in the airport getting set to jet off to London Calling - already spotted a kindred spirit at Logan doing the same, but I could be easily be going somewhere else - or staying home.  There are three major 'omics conferences this week, all in incompatible geographies and overlapping.  Then there is a major vendor announcement day - also in London and perhaps about nanopores and conflicting.  There's also a pub meetup in London that thankfully doesn't  Back in Boston, there's also two free NGS vendor events I might have gone to.  Not only can't I attend them all, I can't attend most - and it will be impossible to monitor Twitter in real time much of the time.

Business travel is always a two-sided coin.  On the one hand I enjoy seeing new sights and revisiting other favorites.  I've been lucky that I've fit some sort of fun into just about every business trip I've taken, even if it's meant riding a nearly empty cable car to a deserted Fisherman's Wharf.  Two exceptions were both day trips, but how I couldn't sneak in excitement at Monmouth Junction New Jersey or East Haven Connecticut won't exactly haunt me.  But I'm also torn about travel: home is where the dog is.  Also the spouse and F1.  

Conferences are also an exhausting rush; London Calling isn't as bad as AGBT in this regard as it is shorter, there's only a late evening one night and there's not afterparties or European friends defeating your well intentioned plans to sleep by bringing a bottle of port and fine Dutch chocolates.  But it's always intense trying to catch talks, make meetings, visit demonstrations and booths, catch up with friends you only see at a specific meeting and take notes to share with colleagues and have a shot at writing something coherent for this space.  I'm not complaining - I could always quit, but I won't anytime soon.  This year London Calling is three days of talks instead of the usual two, so a bit more of the thrilling grind.

And of course,  at least for the next several weeks, there's my current day job to attend to.  Conferences are great and often lead to valuable connections and information; at AGBT i even tried on a Business Development persona which seems to have yielded multiple legitimate leads four our enzyme discovery & engineering business. But in the end, the team I'm on plus myself would like to see progress on the projects I've taken on, and I'm many solar orbits past the age where I can conference all day and code all night.

This year London is also featuring a Diagnostics Day from Roche on Wednesday.  There have been persistent rumors that Roche will finally, over a decade since acquiring Genia and four years after acquiring Stratos, announce a nanopore sequencing platform.  Or maybe not.  But counterprogramming that against London Calling is downright annoying!

If you are in London for either event, the folks at Plasmidosaurus have announced a pub night on Thursday at The Court.   It's over near University College London, which isn't very close to Old Billingsgate where London Calling is, but it appears to be a simple tube ride around the Circle Line. 

There's a list of conferences I'd like to attend but the timing never works; the Biology of Genomes has been "I'll go next year" for two decades.  SFAF is higher on the list, given the location - I love the American West scenery and Santa Fe can be such a great launchpad for so many interesting adventures - plus it sounds like a great conference.  SFAF has a reputation for taking on more of the early stage commercialization technologies that don't have a strong home at AGBT anymore.   Last year it was bumped for the practical concerns of a college graduation, moving the F1 back home and a celebratory trip.  Now it's in conflict with London Calling and put off for another year.

If I'd stayed in Boston, I might have thought about catching  NextFlow Summit.  Understanding workflow languages is becoming critical in bioinformatics (don't ask if I've followed that advice), and NextFlow is one of the leaders.

Back home, PacBio has the Boston edition of their PRISMBoston edition of their PRISM series -- I caught it two years ago and the talks were very good, but last year some schedule conflict of another caught me.  Complete Genomics is having a grand opening celebration at their new "Customer Experience Center" in Framingham, the biomedically biomedically immortal Boston suburb. I won't begin to claim I've checked other geographies; I did see a Nanostring event in central Europe going on this week also.

So I'm off to Gatwick in under an hour.  It horrified my British colleague I mentioned this to - Heathrow does have cachet plus often great views of Windsor Castle on approach - but I pointed out she's the one who got me hooked on picking up delectables at Borough Market, and if you get off the train from Gatwick one stop early at London Bridge the market is almost under the station. I'll get there just around the time it opens - and it's a quite reasonable walk to my hotel near The Tower of London from there.  JetBlue made it even easier last year by sending my bag to Heathrow and then delivering it, but I can't count on such service on a regular basis!

Going to London Calling or the pub outing?  Look for my hat celebrating the birthplace of Taq polymerase and please do say hi! 

HiFi WGS As A (Nearly) Unified Tool For Rare Genetic Disease Diagnosis

What is now way back in February, Alexander Hoischen presented a talk at AGBT which described early results from an effort to apply PacBio HiFi sequencing at scale for solving rare disease cases.  Hoischen passionately made the case for how providing a diagnosis can change affected families.  It's also worth noting how important rare disease genetics has been to the history of biology, illuminating new processes and entire pathways.  Something I hadn't appreciated until his presentation is how many technologies are currently thrown at a case in current workflows because each technology can cover a few types of mutations but miss others.  So this is good snapshot of the current state of human genomics technology with hints of where it might be going.  And Hoischen made a strong case that many other technologies  - but not all of them - can be retired if PacBio HiFi sequencing is the lead approach.  A longer, similar talk is also available as a PacBio-sponsored webinar given by Lissenka Vissers from the same institution and some of the data is in a preprint linked below.

DoveTail Transposes Their Hi-C Methodology

Technologies vying for state-of-the-art in human genome analysis are a recurrent theme in this space, and there are many ideas on this in the collection I really need to get out over the next two weeks before my brain is overwhelmed by London Calling.  Up today: Dovetail Genomics popping back on the scene (as a subsidiary of Cantata Bio)  with an AACR poster several weeks ago showing early results from a "LinkPrep" kit that will commercialize tagmentation (in vitro transposition to fragment DNA and add adapters) for Hi-C library generation, with the promise of enabling short read sequencing to deliver both SNVs as well as long-range structural information all from the same library.  

AQTUAL: Arthritis Drug Selection Via Assaying Cell Free Chromatin

[Note: after I initially released this, Dr. Abdueva spotted some glitches; I pulled it back for editing & then got swept into London Calling; this revised version is finally emerging]

Liquid biopsies - the idea of peering into the disease state somewhere in the body by looking at “cell-free DNA” in the blood - is quite the rage these days.  There are a host of companies and approaches, and I haven’t quite found the discipline to start trying to build a census of all of them.  The field started with Non-Invasive Prenatal Testing (NIPT), and then some early NIPT cases had odd DNA that looked like oncogenic in healthy mothers - who turned out to actually have the cancer. Oncology has been the primary focus, but there’s been many hints that liquid biopsies may be valuable in a wide range of diseases.  A bit over a week ago, Dr. Diana Abdueva founder and CEO of AQTUAL, walked me through (over Zoom) that company’s liquid biopsy approach to inflammatory disease management.

On The Expanding Versatility of Single Molecule Sequencing for Detecting Anomalous DNA

An exciting aspect of true single molecule sequencing has been the detection of methylated bases.  Both Oxford Nanopore and Pacific Biosciences technology generate altered signals if methylated bases are present.  For Oxford Nanopore this is hardly surprising, as it would seem any change in the DNA should alter the complex interaction with the protein pore and it should become just a computational challenge of recognizing that signal.  PacBio is a bit more surprising, but the kinetics of base incorporation are apparently sensitive to the complementary base.  I wanted to point out, though without much deep analysis, three recent preprints that demonstrate detection of other modifications to DNA and thereby enable some interesting applications (and of course, some wild speculations on my part).  It's also interesting because of the overlap between the papers, as they are interconnected to a degree in their methods.

First Illumina Complete Long Reads Preprint

Readers of this space might have detected a significant slant towards skepticism in my coverage of Illumina Complete Long Reads (iCLR), exacerbated by now deposed Illumina CEO Francis deSouza claiming it isn't a synthetic read technology.  Illumina's posters on iCLR at AGBT this year seemed to reinforce my view that Illumina was marketing purely on short-read like terms - call SNPs in a few more hard-to-map regions of the genome, but not really compete head-to-head with the true long read platforms.  But now there is a preprint out on MedRxiv that reports iCLR results for a Genome In A Bottle (GIAB) sample as well as seven samples from individuals wiith potential genetic diseases of unresolved cause.  The GIAB sample was also sequenced with some of the latest Oxford Nanopore chemistry (Duplex R10.4.1) and as HiFi libraries on PacBio Revio - enabling comparisons of the platforms.  The preprint is probably going to be revised and expanded - I'm certainly hoping some of my comments are found constructive - but is very useful to see.  And perhaps it will soften positions such as mine on iCLR's utility.

A Peek At QuantumSI's Protein Sequencer

A number of academic labs and startups have been trying to build new ways of parallel sequencing of large numbers of peptides using schemes that have significant resemblance in their logic to the highly parallel DNA sequencing schemes often highlighted in this space; QuantumSI is the first (and so far only) such company to actually commercialize in this space.  Resemblances to NGS but not identity - for a few important reasons.

The biggest such challenge is the lack of anything resembling Watson-Crick basepairing in proteins. Sequencing chemistries almost invariably rely on basepairing, with the notable exceptions of Maxam-Gilbert reactions and nanopore sequencing.  Even ONT's scheme ends up leveraging basepairing at times, such as the sequencing adapters and various incarnations of double-stranded sequencing (2D, 1D^2, duplex). And very notably, there is not and probably will never be an equivalent of PCR for peptides; any peptide sequencing technology will inherently be a single-molecule approach  

Furthermore, peptide management enzymology just isn't as well developed.  There's some known proteases with degrees of specificity, but nothing like the wide catalog of restriction enzymes you can order from NEB or other vendores.  There's no polymerases of course, but even tools like ligases just don't have as wide a scope - though again, ligation are often driven by some basepairing.  Nature didn't make this space easy!

For these reasons, nearly all of the proposed chemistries are degradative in nature, with nanopore direct reading of peptides making up the rest. N-terminal degradation is an old concept; Edman developed his chemistry around the same time Fred Sanger was first solving the sequence of a protein (insulin) about 70 years ago.  Performing such analysis on single peptides, rather than pools will clearly be challenging - though it does eliminate the phasing problem and the problem of dealing with mixed populations of input peptides such as we did in a paper back yonder.

So the general concept will be to digest proteins into peptides, likely with trypsin, tether those peptides to a solid surface by their C-termini and then progressively read each N-terminal amino acid followed by removal of that terminal amino acid to expose the following one.

One idea for next-gen protein sequencing, with one example pursuer Encodia, is to try to build what is in effect a "reverse translatase" - progressively disassemble a protein and encode the released amino acids as DNA to be sequenced on a high throughput sequencer.  Each amino acid is coded back into DNA using some sort of code words, based on oligo-tagged recognizers.  One challenge with such a concept is the difficulty of distinguishing closely related amino acids, with leucine vs. isoleucine perhaps the most tricky.  The next is that each amino acid must have its own recognizer.  Of course, it might be acceptable to have some compression - maybe isoleucine and leucine aren't distinguished and that is dealt with in downstream search software.  But, even if the amino acid sequence space must, by necessity, be compressed, the total space of interest is huge if common post-translational modifications are desired to be in scope.  And many of these modifications may complicate the selection of recognizers.

QuantumSI is detecting the recognizers directly using optics. Importantly, they are using the time domain as well -- something a reverse encoder probably can never leverage. In fact, they use the time domain two different ways.  

First, each recognizer is labeled with dyes with different fluorescent lifetimes but the same absorbance and emission spectra.  This enables a monochrome optical system, and monochrome is always simpler and higher resolution than a polychromatic system.  Put another way, they've shifted possible optical and/or mechanical complexity into the chemical domain.

Second, the dynamics of the recognizers binding the N-terminus of a peptide are a key part of the signal. Rather than some sort of 1:1 pairing of recognizers to amino acids, each recognizer will display a certain pattern of binding kinetics with each possible terminal amino acid.  QuantumSI says they can distinguish leucine from isoleucine, as they display different kinetic signals. The biggest advantage is that a small number of recognizers can potentially differentiate a very large number of amino acids - QuantumSI's latest chemistry uses just nine recognizers.  They aren't yet claiming decoding all the funky amino acids - from my Millennium life I have not only a love for phosphorylation but also ubiquitination and its kin - but their system may have a shot at many of these without requiring a custom recognizer for each one.

A very interesting design choice from QuantumSI is to make their system a single-pot chemistry; there is no chemical cycling as with their corporate cousin 454.bio.  This makes for a much simpler instrument - a great deal of microfluidic complexity avoided - and saves on reagents since none of the expensive components are lost.  Unlike 454.bio, QuantumSI doesn't even need to remove incorporated labels, since they are degrading the analyzed peptides.  

But, this does complicate things.  There's basically always a race going on for access to the N-terminus of each peptide. Recognizers will come and go, but eventually the N-terminal endopeptidase strides in and clips off an amino acid - and hopefully leaves without clipping another.  In the ideal case, a set of recognizers flit in and out, giving a complex and useful signal, before the clipping - but there's no guarantee of that.  The scheme also seems a nightmare for any homopolymeric stretch - I doubt QuantumSI will be used to count glutamines within huntingtin.  But with looking up in a database, these should be manageable issues -- and the incumbent technique of mass spectrometry has its own challenges.

How simple is the workflow?  QuantumSI says their communications guy ran it.  One hours hands on time to digest the sample and click-label the C-termini for attachment to the flowcell, followed by 10 hours of running.  Automation of this workflow is on their development roadmap.

On the recognizer front, QuantumSI has made steady progress.  Their publication in Science used only three recognizers; at launch they had five and the newest kits have six.  This really emphasizes how their kinetic analysis can extract a great deal of data from a small number of recognizers.  Some post-translational modifications can already be detected, though the high value space of detecting phosphorylation is still in development.

On the informatics site, QuantumSI provides a hierarchy of data, with "what proteins are we identifying" on top, counts of individual peptides the next rung down and detailed kinetic information on each residue at the bottom.  

If QuantumSI is the Answer, What is the Question?

A core challenge with biological mixtures of proteins is the extreme of dynamic range. For example, with human blood (or serum or plasma) you can remove something like 99.99% of serum albumin and the dominant signal will still be serum albumin.  Solve serum albumin and a new set of abundant proteins must be batted down. The really interesting stuff is many orders of magnitude less abundant than all that.  Which is one of the reasons immunoassays such as home pregnancy tests are so amazing - they detect absurdly dilute targets in a sea of abundant proteins yet can be made cheaply and run with essentially no training.  

Some in the mass spec field have been not been shy about pointing out this issue; indeed, some have been downright obnoxious about it. Unless you can sequence enormous numbers of peptides - or figure out some extremely clever ways to deal with those abundant proteins - sequencing approaches will be swamped by boring background.  

QuantumSI's answer to this is to not take on such difficult challenges, at least not yet.  What they are proposing is that m biologists for ages have used tools such as Coomassie staining, Western Blots and ELISAs to study abundant proteins in simplified mixtures, and QuantumSI can provide higher information content but with workflows that are simple to learn and use.  After all, one drawback to mass spectrometry is it requires a very expensive set of instrumentation that requires a high degree of training to operate.  Mass spectrometers with associate liquid chromatographs are not something every lab is going to splurge on; doubly so on the mass spectrometrist to go with it.  QuantumSI claims their sample prep workflow is just a simple set of biochemical steps; no chromatography required if your inputs are simple.

At $85K an instrument, QuantumSI certainly isn't going to be ubiquitous as a simple gel box. Perhaps more seriously, the current instrument processes only two samples at a time, with runtimes of roughly overnight.  That's much less throughput than a simple gel box.  QuantumSI says that for applications so far they are resolving more peptides than required, so expanding the number of samples is high on their priority list.  This also points to another place the nucleic acids have a leg up - it's really easy to design barcoding schemes for DNA or RNA since we can easily design, synthesize and tack on such barcodes; this technology isn't well developed for peptides for direct peptide reading (the mass spectrometrists do have fancy mass-encoded tags).  But there are already case studies using QuantumSI to read out genetically encoded peptide barcodes, so there's already progress there.

Among applications mentioned by QuantumSI: reading out protein-protein interaction partners detected by immunoprecipitation, verifying protein engineering results, quality control for antibody production., and verifying if an engineered protein mutation is being correctly expressed.  All applications where the number of abundant proteins is sufficiently low to avoid the signal of interest being swamped out.

QuantumSI commented on the sorts of conferences they've attended and the response.  The Festival of Genomics - I first saw a box in the wild at FOG Boston last autumn - has been very successful, as has been other genomics-oriented conferences.  In their view, genomics practitioners are reluctant to invest in mass spectrometers.  They also go to proteomics-oriented conferences and encounter a much more mass spec oriented audience and the skepticism for NGS-like approaches held by that community.  Currently they are selling themselves in North America and Europe and using distributors to sell into Asia-Pacific geography.

It will be interesting to watch the further development of this space.  QuantumSI launched at the end of 2022 and is still the only NGS-like protein sequencing that has launched.  The new kits just announced have increased the number of peptides read out by about two to seven fold.  Personally, I think having more sample chambers per run is likely to be very popular; nobody ever ran a two lane gel!  And it may take time to identify the "killer apps" which will drive labs to buy into the platform, though even a few splashy publications could create some significant buzz.  

A final thought: it's interesting that QuantumSI gets attention at genomics-oriented meetings, but how much low-complexity protein sequencing are genome-focused labs interested in?  Perhaps it is a new direction that some are contemplating branching out in, but in general I don't see the QuantumSI approach - at its current level of sample throughput or tolerance for sample dynamic range - being a frequent companion for high throughput genome sequencing, RNA-Seq or spatial analysis.  There is an apparent fit for smaller scale synthetic biology and protein engineering labs perhaps - it remains to be seen how many such labs will try this technology out.  Rather than core labs, I suspect the better fit for QuantumSI is individual principal investigators or their equivalent in industry.  That is a very diffuse market with weaker network effects to drive adoption (versus genome labs that love to get on the latest bandwagon).

Bruker Wins NanoString Auction

NanoString declaring bankruptcy on the eve of 2024's edition of AGBT was a shock to many at the meeting and then there was confusion: would one of the sponsors have a dark booth? The aggressive 10X Genomics legal strategy that forced the bankruptcy raised a degree of polite ire. But NanoString marketing carried on and CSO Joe Beecham delivered a fiery speech saying "we're not going anywhere". Then an investment firm, Patient Square Capital, appeared to be the front runner for acquiring the assets, with speculation they would combine NanoString with their other spatial omics portfolio company, Resolve Biosciences.  But last week, as the genomics world was still processing PacBio's turmoil, news broke that Bruker had significantly outbid Patient Square - $392.6M vs $220M.  So Bruker takes NanoString home - and I gives me an entree to float an ontology of spatial technologies I've been fermenting, as Bruker will now have instruments in the four major spatial approaches.  And 10X now has a more formidable opponent in the ongoing patent wars.

PacBio Plummets

PacBio announced preliminary earnings yesterday, and the nearly immediate result was a 50% plunge in their share price.  Along with the earnings, the company announced significant cost cutting.  The details of those cuts were not made available, but some clever tea leave parsers noted a significant omission from what the company said it would continue.  The ASeq Discord channel on PacBio absolutely blew up, with opinions ranging from PacBio is in a death spiral to PacBio must be for sale, with significant numbers of "Christian Henry won't be CEO by year's end".  

Thoughts on RNU4-2 Mutation Paper

A new preprint based on Genomics UK data has identified a set of single base insertion mutations (predominantly a specific A insertion)  in a spliceosomal RNA which is responsible for about 0.5% of previously undiagnosed genetic cases of syndromic neurodevelopmental disorders . That's a remarkably high frequency mutation which has gone unnoticed to date, but the fact it was hiding in a non-protein-coding RNA (a spliceosome component called RNU4-2) had much to do with that - this gene won't be in any exome panels. The mutation always appears to be de novo and therefore the pathogenic phenotype is dominant.   I'd like to write down a few other thoughts - mostly in the form of questions --  with the caveat that I've never worked on a rare disease project and to describe me as a detached armchair voyeur of the field would be far too generous.

Post-AGBT: VizGen & Scale Biosciences Partner

It's been just a few weeks since I sat poolside at AGBT with VizGen CEO Terry Lo and Scale Biosciences CEO Giovanna Prout to discuss the two companies' new partnership.  Well, that would have been accurate about a month ago; getting the last AGBT threads together has been buried under post-AGBT day work, some family business, another vacation - and let's be serious, mega-scale procrastination and writer's block (and that's just a euphemism here for more procrastination).  But that shouldn't detract from what these two RNA (and more!) profiling companies are trying to build together.  Plus this is my last "Post-AGBT" tag for the year; now I can move on to "inspired by AGBT" that is a bit less tied to the meeting (and less obviously overdue)

BioNano In Peril Again

While I still have a pair of pre-AGBT and AGBT interviews to write up - plus a long list of post ideas inspired by AGBT - breaking news about BioNano Genomics takes precedence.  The company has announced a major restructuring, with about 30% of its employees being laid off.  I've been laid off twice and it's never enjoyable, so I hope what I write here is appropriately sensitive - but won't be surprised if I still commit a faux pas.  Even with the restructuring, one analyst who likes BioNano estimated they will have about three quarters of cash - this is indeed a perilous time.

Post-AGBT: Sequencing Hardware Roundup

Some updates on the sequencing instrument vendors, save Ultima Genomics and Element Biosciences which I've covered already.

Post-AGBT: Element AVITI Sequencing Updates

Element has been very busy over the past year and in the Silver Sponsor presentation covered updates since last AGBT as well as a number of completely new items.  I covered their Teton approach to multiomic analysis of cell culture in the last piece; in this one I'll cover their sequencing platform evolution.  Element was kind enough to loan me key members of their technical braintrust for an hour in the week before AGBT, which sadly I repaid by allowing their lunch to be scheduled over.  Thankfully, they do have a recording available!

Post-AGBT: Both Element & Singular Want Spatial to Go With The Flow(cells)

Element Biosciences and Singular Genomics have often appeared to be on roughly parallel trajectories, though with key differences.  Both companies launched sequencing instruments with NextSeq 2000-like specifications and largely aimed at the academic core lab and small biotech company market.  At AGBT, both announced upgrades to their sequencing instruments that allow the instrument to perform spatial omics while still functioning as a sequencer.  But there are key differences in their approach and what we know about each company and their degree of success so far in the sequencer market.

AGBT Follow-up: Ultima Genomics UG100, Volta Labs Callisto, N6Tec iconPCR

A confusion of ideas for AGBT follow-up have collided with the inevitable post-AGBT return-to-ordinary-life requirements.  To try to avoid a huge project that never gets completed, I'm breaking these up into multiple pieces.  First off, a look at reaction to the three big pieces I wrote before the conference or early during the conference: Ultima Genomics, Volta Labs Callisto and N6Tec iconPCR.  My comments are based on further thoughts on my part, discussions with other AGBT attendees and feedback I've gotten via social media, blog comments and emails/DMs.  Please keep it coming!  One of the great values of writing this is getting feedback - it illuminates questions I haven't considered and highlights gaps in my thinking. 

VoltaLabs Launches Callisto for DNA Extraction & Library Prep

Here at AGBT, VoltaLabs has unveiled their 24-sample DNA extraction and NGS library prep Callisto instrument, which is particularly suited for long read applications but is also suited for short read work. Volta has matured liquid handling automation to a novel open top electrowetting technology. Priced at $125K and planning to ship in the second quarter, Callisto is designed as a walk-away solution requiring no human interaction during a run. Personally, not only do I love the a new medium-throughput instrument for HMW DNA extraction and manipulation, but I also can at least pretend I helped steer the company In that directions

iconPCR: Super-Flexible qPCR Thermocycler Oft Dreamed, Now Delivered

Has there ever been a product you’ve just wanted to have, but it doesn’t exist? That keeps popping up in discussions - “if only we had X this project would go so much faster!”. Well, N6 Tec’s automation-friendly $99K i96 well iconPCR thermocycler is that to me. Launching at AGBT, it’s the gadget I’ve wanted repeatedly at Codon Devices, Warp Drive Bio and now Ginkgo Bioworks. It won’t solve all your PCR challenges, but it certainly gives new options to customize PCR like never before. And for many NGS labs, it offers major streamlining of PCR-based library construction protocols while also delivering superior data. How? By being a thermocycler where every well can run its own thermal profile and each well can go dormant once a desired level of amplification is achieved 

Want to Build A Sequencer? 454.bio Opens Up Their Plans

Just as the AGBT hype cycle was firing up (with me contributing multiple sparks), serial entrepreneur Jonathan Rothberg's latest sequencing startup 454.bio fully de-stealthed their technology this weekend, going so far as to release open source plans to build an instrument prototype.  454.bio  is aiming to build a Keurig-sized device to retail for $100, with sequencing runs in the $20 range.  To accomplish this, they're attempting a novel twist on sequencing-by-synthesis.  It's an unconventional strategy by someone who has succeeded twice before in DNA sequencing (454 and Ion Torrent) and has multiple other companies going (if I've counted correctly)  - QuantumSI in protein sequencing (a future topic for this space, I promise!), ButterflyNetworks with inexpensive, compact diagnostics ultrasound and Hyperfine with inexpensive, compact MRI diagnostic devices.  Then I went to the 4Catalyzer site - Rothberg's incubator - and discovered a bunch of companies I hadn't heard of or had forgotten about -- Protein Evolution in synthetic biology for plastics production, Detect for home-based diagnostics instruments, AI Therapeutics in the rare disease space and Liminal with what looks like consumer brain scanning.  That's quite a series of companies!   But the one closest to my heart (sorry QuantumSI :-) is  454.bio, and their announcements have many interesting facets which I'll dive into.

[2024-02-06 01:41 - 'used"--> iSeq fix -- stupid autocorrect!]

Ultima Launches

As part of the run-up to Gold sponsorship at AGBT, Ultima Genomics held a multi-day event in early December, with tours of the headquarters facility and factory floor in the Bay Area and a day at a beautiful Wine Country resort. The resort session included talks from the company, early access collaborators and a pair of big name early backers, with a few hundred current customers and many contemplating the leap.  So confident was the company in their product, they even invited a blogger to moderate one of the panel discussions!  The UG100 is now officially launched as a fully commercial product, with ambitions to replace panels, exomes and microarrays with whole genome sequences at $100 apiece.  All in an instrument package designed for continuous industrial-scale operation.  Please note that Ultima did review this piece to ensure I didn’t disclose information they did not wish public, but for the most part just gave me some very good proofreading support.  Photos are my own, except as noted.

On Illumina's Moats Past & Present

Studying how Illumina came to dominate sequencing markets is certainly worthy of at least a Harvard Business School case study, and perhaps an entire graduate thesis.  But I wanted to give a quick review of some of my thoughts on the matter, spurred by Nava Whiteford's repeated savaging of a piece in another space but also because many of these themes will show up in a flurry of pieces I'm planning (one's even nearly done!) in the next few weeks due to AGBT and some non-AGBT news.  

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!