Thursday, February 20, 2020

A Lazy Look at The Field of Sequencing Startups

AGBT looms ahead of me next week which serves as impetus to let fly an idea I've had simmering for a while: to look at sequencing startups by a particular type of information they choose to reveal.  I'm not expecting any big announcements at AGBT from this space, though would be thrilled to be surprised.  But there is the risk of getting contaminated with some on-the-sly scuttlebutt, so better to get this done now.  By the way, in the full disclosure category, I have consulted for a few companies here and have NDAs either on my own or via employers; everything here is based on public information.
Many startups operate in some degree of stealth.  That has many advantages.  First, one wants to limit the information competitors have to beat you or thwart you.  It also adds to the mystique which can be dramatically tossed away at a big reveal.  Conversely, stealth mode can be a hindrance to recruiting and prevents potential collaborators.  So management types often think carefully about what information to make public.  Not that they always get it right; Genapsys seemed in 2014 at AGBT to be signalling they were near launch, even recruiting for an early access "Genius Club", but wouldn't launch until last fall.

There are various types of information that companies do reveal.  There can be scientific talks or posters.  Those are often hard to cover unless you're there.  There's giving interviews to reporters and bloggers.  I'm happy to help there; perhaps AGBT will stimulate that route.  There's also patent filings; those are real work to analyze.  But for someone looking for a quick-and-lazy route, there is going to the websites.  Now, that is one of the most controllable; Starbase's was gloriously uninformative while we were staying cloaked.  But as a company gets closer to launch, the value of releasing information -- particularly for recruiting -- will grow while the value of stealth should be fading away.  

So what I will do here is report some notes on just looking at the websites of startups in the general purpose sequencer field.  So that rules out anyone now in the fold of a large company, such as Genia (Roche) or LaserGen (Agilent).  It also rules out someone like ReadCoor that appears to be aiming for a very specialized market of spatial sequencing.  To accumulate a list of targets, I've relied on the many notes in Nava Whiteford's 41J blog and Albert Vilella's wonderful compilations.  Plus one company had a splashy financing after I started this.  I've broken them into several categories,  both because I have that urge nearly reflexively as well as the list is long

Stealthy? Or Dead?

One challenge with tracking startups is since they aren't really obligated to make noise, it's really hard to tell when they have ceased operations.  If you are truly stealthy, then even a website may be too much -- and if you've given up then a website is money down the drain.  

For example, ZS Genetics is probably defunct; I certainly don't see an expensive electron microscope platform co-existing with the bevy of inexpensive devices that are rolling out.  But I actually pass infrequently the business park where they were last located and the sign is still on the outer signboard -- though that could also reflect that the space isn't in high demand.

Others of which I could find no electronic home are Cygnus Biosciences, Esper Biosciences and Eve Biomedical.  Genome Surveillance has a locked pitch deck online.

A special comment must be made about ElectroSeq.  No website per se, but they do have a Facebook page.  More curiously, there's a mention elsewhere that not only are they developing new sequencing technology, but they also provide services in the area of horse genetics.  That's perhaps a creative means of early stage funding, but leads to the next category...

Probably Too Distracted

It's not impossible to launch a new sequencer as part of an existing business, but it's generally going to be very difficult.  Even if some other business is a cash cow, there's always the danger of a lack of management focus on the blue sky activity.  Sequencing was in the original business plans for both OpGen and BioNano Genomics, but the mapping business took over -- plus the rest of the market likely outpaced much of what they were planning.  Nabsys had a similar mapping as prelude to sequencing aspirations; it's hard to see the mapping business giving them any boost to get into sequencing.

I profiled Universal Sequencing's TELL-Seq method recently.  That's in the right space and there may be synergies, but ultimately developing kits and developing hardware are different beasts.

Ontera, formerly Two Pore Guys, I also mentioned recently in conjunction with the COVID-19 outbreak and diagnostics.  Their technology has very interesting possibilities for portable rapid detection of a wide range of analytes, but that in my opinion will mostly steer them away from sequencing - the data requirements are just radically different.  Similarly DNAe over in England has rapid diagnostics, but those aren't sequencing.  AxBio in China is another company working on electronics for rapid diagnostics; it's all too easy to see success in diagnostics forever pushing sequencing to the back burner.

Centrillion Technologies is another curious case.  They're a sequencing service provider which frequently is mentioned in both the DNA sequencing and DNA synthesis areas and patents pop up for them.  But go to their website and nary a mention of any of this.

We Exist and That's All You Need To Know

If you want something barely above stealth, check out XGenomes website -- it gives you their physical location and essentially nothing else. Singular Genomics says "we're busy in the lab". Quantum SI has only some slogans and a warning to contract recruiters.

Molecular Research Technologies aka Mobious is a somewhat notorious outfit.  Back in the days of The Biotech Rumor Mill, that high noise site would frequently see claims that Mobious had a revolutionary sequencing technology around the corner -- none of which ever came to anything.  They appear to have the same stock DNA photo as website background as several other companies.

Patience Please?

There are a large number of companies with websites that actually have some information, but nothing that would suggest they think they will be entering the market anytime soon.

For example, iNanoBio has a glossy page describing the team, but all of their technology photos appear to be very generic. Caerus MDx has no photos but a single sentence that at least tells us it is a single molecule technique.  

Apton Biosystems will be using super-resolution microscopy and "low cost" flowcells; beyond that there is little detailed. 

Armonica Technologies has a scanning electron micrograph which may or may not be related to their "tortuous nanopore" technology, which they claim can sequence without any library prep.

Depixus (formerly PicoSeq) touts that their technology will read the over 20 modifications of DNA and more than 120 reported for RNA. But fewer details of their "MAGNA" technology are given here than at the AGBT talk they gave (2013?).  

Geneseque has a number of details on their chemistry, which is unusual and appears to have some aspects of sequencing-by-ligation.  Detection is via watching where tiny beads are; if still ligated to the target they will be at one position whereas if they are released they bob up from magnetic forces.

Northshorebio has a nanopore sequencing-via-exonuclease scheme.  They are at risk in the distraction area, boasting their technology can be used for non-sequencing applications such as biothreat detection. They have a video, but it's all animation.

Quantapore has a scheme for strand sequencing with nanopores -- their diagram could have fallen from an Oxford Nanopore slide deck -- but with optical detection rather than electrical.  Sample prep is claimed to be only half an hour and they claim they will deliver a $100 genome (presumably human) with reads "on the order of kilobases" in length.  They also claim their cartridge will support from "hundreds to hundreds of thousands of nanopores".

Single Technologies is a company hard to place.  They burst onto the scene very recently sounding more like a company looking to license their super-resolution imaging and fluidics, but on their website they talk of a "Theta" sequencer as well as a spatial sequencing instrument.  

GeneMind in China appears to be the successor to Direct Genomics, which was trying to re-launch the Helicos technology in a desktop format. It's a bit unclear from their English language website where things stand; one press release touts a planned 2017 launch of the instrument.  Most of the papers on the publications page, if not all, were actually run on Helicos boxes.

Element Biosciences burst on the scene earlier this year with $100M in financing and the leadership of Molly He, who once led Illumina's protein engineering group.  They promise they will "retool all of the fundamental elements of a DNA sequencing system: surface chemistry, sequencing chemistry, detection, and data analysis."  But no details beyond that.

Maybe A Year or So Out?

In this category we see more signs of getting close, such as what appear to actual images of system components.  Now, these are educated guesses so certainly this and the previous category probably could have some swaps.

Base4 boasts they will have microfluidics and are single molecule -- they even claim to manipulate single molecules.  Plus it will read out methylation. They have a very detailed overview of the biochemistry, even going into detail (with diagrams!) of how they will capture individual DNA molecules onto magnetic beads. Then their "sequencing by de-synthesis" (clever!) reaction occurs, with each base lopped off via pyrophosphorylysis captured in a droplet, in which an amplification reaction occurs to generate the final signal which is read optically.  That level of detail might suggest getting close to launch, but the few open positions are all technical -- though you really should visit the page for the striking image of a woman in headdress and mask makeup looking ready to stab something with a micropipettor.

Electronic Biosciences' site has images of what appear to be real lab work, not stock photos -- things look very breadboardy but real.  But perhaps this should be in the prior group; there's not much more detail.  It also isn't obvious they are focused on sequencing.

Quantum Biosystems claims "the world's first commercial quantum sequencing systems".  That's a bit of a stretch, given that fluorescence is inherently a quantum phenomenon.  However, they actually have plots on their site which would appear to be claiming to be real conductance data. There's a bunch of images that I think are electron micrographs and not animations, though it's hard to be certain.  

Insilixa has a lot of images of flowcells using CMOS technology.  But it isn't clear that they are really trying to be a sequencing company but perhaps more of an supplier to various companies in the space.  Another one in the potentially distracted field, given their website touts many uses for the technology. An interesting claim is that the flowcell chips can be thermocycled

Roswell Biotechnologies has images of their ENDSeq instrument and of flowcells.  Beyond "direct electronic sensing", we don't learn much of the chemistry.  The instrument photo is the one a colleague was skeptical of whether it was a computer rendering, but at least one chip photo has hard-to-fake human fingers in it. Job listings are limited to a hardware engineer.

Stratos Genomics has images showing a sequencer with readable status information.  It's nanopore based, but if that front page is to be believed not particularly long reads -- it has a mean read length showing of only 450+ bases but with 92+% of Q30 or better.  This is nanopore based detection of "expandomers", bases with wickedly huge modifications which can be chemically modified to create very distinguishable targets for the nanopore.  The template prep is claimed to be 60 minutes wallclock, 15 minutes hands on.  They also show what they say is data off the instrument.  But all the open positions are in technical fields; no commercial ones.

Closest to Breaking the Ribbon?

Omniome is probably the best bet for a near-term launch, as not only is their website quite detailed but they also have job postings for a variety of reagent manufacturing and commercial organization positions.  Curiously, they don't have a prominent technology explainer.  But it is that hiring of commercial positions -- a scientist for "applications and collaborations" and a "senior business development manager marketing/commercial" which are suggestive.

Closing Remarks

Did I miss anyone?  Did I miss any blatant clues that would lead to classifying one of these companies differently? As always, feel free to leave a comment, email me, DM me on Twitter or catch me at AGBT.  Or if you are in Boston, ping me and I'm happy to chat sequencing over noshes at the delicious Flour Bakery.


Danilo said...

Excellent summary - and very educational for someone like myself who is not from the field. A couple months ago I would have stated with utter confidence that nobody is dreaming of launching new sequencers with Illumina and MGI out there. Shows again that one should be careful making statements about things one knows nothing about... But what I'm missing in your post is Genapsys. You mention them at the beginning as a bad example but now they really seem to be ready to go to the market. At least we had a dummy machine from them last week for a show we did in Switzerland. Feedback we received was somewhat mixed and I was hoping you'd give us your thoughts on that one.

Anonymous said...

Like you, the clues I use for the same type of evaluation are:
1) Do they have a commercial person on their leadership team?
2) Are they hiring commercial and operations people?
3) What is their funding history? Launching a new sequencer is difficult if you haven't raised >$100M (probably more like $200M).
4) Any recent news/events? Stratos has a nice shiny website but they haven't made blog or news updates in years.

I suspect that most of your "year out" companies are either more than a year out or struggling. I would move the "year out" companies to "patience please" and move OmniOme to 1-1.5 years out.

Anonymous said...

My prediction is that none of the "year or so out" companies 3ill have a product out within an year.

James@cancer said...

A great post Keith. I hope you're enjoying AGBT and the 20 year anniversary parties (WIWT)?

Considering that most of the cash being pumped into these companies is coming off the back of Solexa/Illumina's successes and that the original GA was launched almost 15 years ago, do you see any slowdown in investment and spinout of new tech? Has it got harder? Will it ever?

Lastly, at AGBT in 2014 an investor spoke to me about my April fools post on the Norfolk Nanopore company - they suggested that with a little more "work" I could have raised a few millions; I am sure between the two of us we can come up with something worth spinning out!

Unknown said...

To know more about the startups and latest genetic information. Please look onto the link below.