41J Blog

In this blog post I’m going to briefly review Chinese DNA sequencing start up Cygnus Biosciences. If you’re interest in other DNA Sequencing startups, check out my list of startups, which also links to relevant summary blog posts.

I enjoyed reading about the Cygnus approach and the error correcting scheme is entertaining for a comp. sci. like me. If you’re into information theory I recommend skipping past the business and chemistry sections, or just take a look at their 2017 paper [2] which describes the method in more detail (and has lots of cool figures).

Business

Cygnus Biosciences is a Beijing company founded in 2017. Reports state they have raise a series B round. However I can only find evidence of a single round taking place. Perhaps there was earlier grant funding, and due to its large size the 19.6M USD round is taken as a series B, it’s unclear.

Investors include Proxima Ventures Ltd, Beijing Longpan Investment Management Consulting Center, Zhongguancun Development Group, Shanghai Creation Investment Management Co Ltd, General Technology Venture Capital, and Beijing Jinggong Hongyuan Venture Capital Center [3]. All investors aside from Proxima appear to be based in China.

Technology

Cygnus appear to have two novel technologies at play. These are both covered in a single international patent [1], which claims priority to a few other Chinese patents (which I’ve not tried to track down). The patent is huge at 224 pages. It contains a lot of data, a lot of figures (pictures of wetting angles, schematics and graphs galore). I’ve not read it in any depth. The basic mechanisms seem to be more clearly described in their papers.

Sequencing Chemistry

The sequencing chemistry was originally described in a 2011 paper from Harvard University [5].

In this system, non-fluorescent bases are introduced into the system. When these modified bases incorporate a fluorophore is released into solution and can be detected. Because the fluorophores are released into solution, the reaction and detection needs to take place in a sealed well with DNA immobilized on beads. The intensity of the fluorescence increases with the number of bases incorporated. As with all such un-terminated technologies there are likely issues with longer homopolymer runs.

Overall, this seems somewhat similar to 454s technology (which ultimately didn’t fare so well). The key difference here is the detection method. On 454s platform the signal was detected via a luciferase-catalyzed reaction. This reaction produced flashes of light that were detected by a camera. The wells therefore had to be monitored during incorporation. This meant it was impossible to have multiple imaging regions, which ultimately made it difficult to scale the throughput of the platform. Other platforms (Illumina) scaled better, and out-competed 454.

With the method presented by Cygnus, you can run the incorporation reaction and then detect the fluorescence at a later point. This means that rather than being limited by the size of your camera, you can just increase the chip size and scan the camera across it to increase throughput.

The original paper shows reads of around 30bp. The 2017 paper describing their error correction system shows much longer reads (200bp+). There isn’t much on the chemistry in the main paper, so it’s not clear what changes have been made. For those interested, it may be worth digging into the patent and supplementary info.

Error Correction Method

Cygnus also present an error correction method. The method can actually be applied to any un-terminated single channel SBS chemistry (Ion torrent, 454 etc.). Rather than detecting the incorporation of a single base type, the Cygnus system uses mixtures of multiple base types (as such it may be somewhat related to the Centrillion method). They then sequence the same template  multiple times using different base mixes before computationally combining the reads to resolve the template sequence.

It’s a neat system, and it’s well illustrated in the paper:

You might well just want to refer to the description in their paper. But for my own benefit I’ve been playing with the method myself.

Conceptually we can imagine generating all possible unordered groups of 2 bases (AC, AG, AT, GT, CT and CG). These are going to be the base mixtures which will be used in the sequencing process.

To build our sequencing system using these mixtures we will obviously need more than one type of mixture. But let’s start with one. A mixture of G and T bases. We’ll call this the “K” group [4]. What will happen in the degenerate case where we have a AC repeat followed by a TG repeat i.e. ACACACACTGTGTGTG? Our read system will just tell us we incorporated the “K” mixture 8 times and then stop because there are no A or Cs in our mixture. So to process the whole strand we need to add second mixture, a mixture of A and C bases we’ll call this mixture the “M” group.

If we alternate between incorporating “K”s and “M”s our read system will tell us we have 8 Ks followed by 8 Ms. Not very useful as a sequencing platform…

In order to resolve this ambiguous read we add read the same strand of DNA again, but this time use a different mixture of bases. Which groups? Well we want to be able to determine which of the two “K” bases was incorporated and which of the two “M” bases was incorporated.

There are actually two different ways of doing this. We can either use alternating mixtures of A/T and C/G (let’s call them W and S groups) or mixtures of A/G and C/T (R and Y). These two alternatives are illustrated in the Venn diagrams below:

Let’s say we use the W and S groups. So we have one read, where we alternated incorporating K and M mixtures, and another where we alternated incorporating W and S mixtures. Our template was ACACACACTGTGTGTG. So the KM read is: KKKKKKKKMMMMMMMM. The WS read is: WSWSWSWSWSWSWSWS. Now let’s deconvolve the read.

Our first base is in groups K and W, only G is in both groups K and W, so our first basecall is a C. Our second base is in groups K and S, only G is in groups K and S so must have incorporated a G and we call a T. Proceeding in this manor we can uniquely identify the base at each position in the template.

So with two reads using alternating mixtures of two different types we can completely resolve the original template. Neat! We’ve used more sequencing, but only two more incorporations (more about this later).

But wait! What was the point? We’ve just sequencing the template, we could have done that without any of this mixture stuff. Well…why stop at 2 sub-reads? Why not add a third and see if that provides any extra value? Remember we had two options in our Venn diagram above for deconvolving the original read? We can add the other 2 groups into our Venn diagram and create this crazy thing:

OK, so now you’re reading each template 3 times. There is obviously redundant information. So if your reads indicated that the first base was in groups “TG” “CT” and “AT” you only need two out of three groups to know that T is the only common member, and that you should call and A. But it your reads indicate that the first base was in groups “TG” “CT” and “GC” clearly something is wrong. There are two groups with each of T,G and C in them. We can’t resolve the base at this position, but we do know something went wrong and can call an ambiguity (“N” base) at this point, preventing the error from effecting downstream analysis.

In general however, errors are unlikely to be of this type. What’s more likely is that a run length is likely to be miscalled. So 3 “W”s in a row will be called instead of 4. These types of errors can not just be detected under this encoding, but also corrected.

The paper claims that they can reduce error rates on their platform from 0.96% to 0.33%. I mentioned previously that more sequencing was required. However, this is only true in the degenerate case. In a normal single channel SBS system you don’t generally incorporate a base every cycle (or flow in Ion torrent jargon). In this system you always do and on average the paper suggests that say 2 bases are incorporated per cycle as opposed to 0.67 bases on other platforms. However, there are clearly degenerate cases (or low complexity cases) where this is not true so the system relies on the template being more or less random (I’d be curious to know what would happen in a high CG content template for example).

Overall, the error correction system seems like a neat idea. It will be interesting to see how it compares to reversible-terminate chemistries like Illumina’s in practice.

Notes

[1] Patent:
https://patentscope.wipo.int/search/ko/detail.jsf;jsessionid=E1CFF0A5F71C9CCAF85848F1A54CA650.wapp1nA?docId=WO2017084580&recNum=692&office=&queryString=&prevFilter=&sortOption=%EA%B3%B5%EA%B0%9C%EC%9D%BC%28%EB%82%B4%EB%A6%BC%EC%B0%A8%EC%88%9C%29&maxRec=3070432

Click to access WO2017084580A1.pdf

[2] Nature Biotechnology Paper: https://bernstein.harvard.edu/papers/2017_Chen_NBT.pdf

[3] https://www.crunchbase.com/organization/cygnus-biosciences#section-funding-rounds

[4] These are the group identifiers used in the IUPAC ambiguity codes: http://www.dnabaser.com/articles/IUPAC%20ambiguity%20codes.html

[5] https://bernstein.harvard.edu/papers/NAT%20METHODS%20v8%20575-80%20JUL%202011%20SIMS.pdf

Business

Centrillion was founded in 2009 by ex-Affymetrix staff. Of the current leadership team only Wei Zhou (CEO) was at Centrillion at its inception and previously worked at Affymetrix. Wei Zhou worked at Affy as an SVP and left after 8 years in 2008 according to LinkedIn [1].

Centrillion has a more complex corporate structure than I’m used to seeing in a startup. The governing entity for Centrillion appears to be Centrillion Technology Holdings Corporation which is a Grand Cayman corporation [9]. Centrillion Biosciences Inc. [2] is a subsidiary of the holding company, and is located in Mountain View.

They appear to have offices in Palo Alto, Portland, Hangzhou-China and Hsinchu-Taiwan [3]. They mention that they acquired “assets” from Perlegen and Affymetrix, but it’s not very clear what these assets are. Given that they also tried to acquired Affymetrix in 2016 I would guess that there was IP and staff and other assets that they were interested in acquiring at that point (this brought another company into play, Origin Technologies Corporation, LLC) [5]. Affy did not accept the Origin offer, suggesting that the ThermoFisher bid was more realistic [8].

Centrillion seems to be funded mainly by Chinese investors. SummitView Captial [4] is among these, and also backed their Affymetrix bid.

Centrillion has launched at least one service, Tribecode. This seems to be a genetic genealogy testing service. Reports suggest is uses low coverage Illumina sequencing. At present the tribecode website (www.tribecode.com) points to an unconfigured webserver. The service is however mentioned on the Centrillion website. It’s unclear if this is a temporary issue (it’s been offline for at least a couple of days) or if the service is no longer available.

Glassdoor reviews for Centrillion are interesting [17], many reviews complain of cameras monitoring staff work areas and being forced to clock in and out of the office (and reprimands for being in the office less than 8 hours). One review notes “Company appears to be a front for sheltering Chinese capital investors rather than an actual product or technology driven enterprise.”. I tend to take glassdoor reviews with a grain of salt, but it’s interesting gossip.

Last week (19th July 2018) Centrillion are reported as raising 58.78M USD [10] [11]. In total they appear to have raised in excess of 86.79M USD [11]. In addition to this, they appear to have received 650,000USD in SBIR grant funding [12].

Technology

The Centrillion website has a services link, this currently points to services.centrillionbio.com which I can not access currently (timeout). Looking at archived copies of the site at the Internet archive suggests that all services are carried out on Illumina instruments. In particular they reference Hiseq2000 series instruments (which are now quite old): “sequencing to desired coverage using Illumina HiSeq2000 or HiSeq2500 instrument”.

Centrillion also runs “Molecular Vision Lab”. This subsidiary seems to have a number of panels available for vision related genetic disordered. The sites states “This new test is a capture based panel with baits designed and manufactured by our parent company, Centrillion Technologies.” [13].

None of these services sound quite like the statements from Centrillion’s factsheet which suggest development of a DNA sequencing platform:

“Centrillion is leveraging an extensive intellectual property portfolio and the latest advances in engineering, chemistry and biochemistry to develop an array of solutions, from targeted sequencing to chromosome scale sequencing; from genotyping to 3D gene expression, all based upon a unifying genomic analysis technology.

Centrillion’s work is supported by a new generation of sequencing-compatible DNA chips, which combine the high density of first-generation chips with the high-quality probes of more recent innovations, all in one affordable and scalable mode.” [16]

So I’ve tried digging a bit further to find more details of sequencing technologies they might be building.

One patent, “Methods and Systems for Sequencing Long Nucleic Acids” [14], discusses a modification of SBS protocols to enable longer read sequencing. The method relies on the use of existing SBS IP and discusses demonstrations on Illumina and Ion Torrent platforms. Essentially they suggest performing multiple reads against the same clusters, at offset positions.

So you would perform a normal single sequencing read. Then remove the synthesized strand. Then re-prime the clusters. This essentially resets the sequencing experiment. You then use one of a number of methods to advance the start position. A few methods are discussed. Most of these seem to rely on what they call “limited extension”, where they incorporate bases but limit the number incorporated somehow e.g. by

* Limiting the amount of time.
* Leaving out one base, so the extension stops at that base.
* Having one base use a reversible terminator to stop the extension.

The examples they present use what they call +S, dark base extension. Dark bases are just native nucleotides, i.e. unlabeled. On an Illumina instrument after “resetting” the clusters they perform a number of cycles with mixtures of these dark bases. For example you might first do a cycle containing A,T and G but not C. This would advance the cluster to the next C position. Then you wash, and do another cycle with C,A,T but no G. A number of these cycles can be performed to advance the read position. After you’ve completed this “limited extension” process you can perform your read as normal.

Neat idea, there’s some data in the patent but I haven’t taken the time to dig into it.

There’s another patent “Native Extension Parallel Sequencing” [15] which also seems to use the limited extension concept. In this patent, a capture system is used. Using this you can have multiple spots/regions which contain the same template. The patent then uses a system which to be frank I don’t quite understand, to obtain partial information from each spot which is assembled into a complete read. It seems interesting, and if I had time I’d like to dig into this patent more and better understand it.

These are both quite old patents… more recently a paper co-authored by Centrillion staff has been published. This paper is discusses on surface library construction. Again this looks interesting, but I’m more interested in sequencing technologies.

There are a few other bits and bobs like this patent for a reversible terminator. But nothing that indicates to me serious development of an SBS platform. So I’m still curious as to what they are actually developing. It seems possible that there interests have shifted from sequencing toward prep/capture/library construction and running a services business.

Notes

[1] Leadership team from the Centrillion factsheet, who worked at Affy. Durations and dates from LinkedIn:

Wei Zhou (CEO) was SVP at Affymetrix until 2008 (>8 years)
Janet Warrington, VP R&D until 2008 (12 years). Joined Centrillion 2014.
Glenn McGall, VP; Research and Development, Chemistry until 2012 (5 years)
Suzanne Dee, Director, Expression Product Development, Jul 2014 (>18 years)

[2] https://www.bloomberg.com/research/stocks/private/snapshot.asp?privcapId=72011751

[3] Office locations from job posting, https://startup.jobs/62911-research-scientist-at-centrillion
“We have advanced research laboratories in Palo Alto, CA; Portland, OR; Hangzhou, China and Hsinchu, Taiwan”

[4] http://en.summitviewcapital.com/plus/list.php?tid=10
SummitView Capital backed acquisition: https://cen.acs.org/articles/94/i13/Former-employees-seek-acquire-Affymetrix.html

[5] https://www.sec.gov/Archives/edgar/data/913077/000161577416004576/s102858_dfan14a.htm
Mar. 27, 2016: https://seekingalpha.com/article/3961150-week-review-origin-centrillion-offers-1_5-billion-affymetrix
Centrillion fail to buy Affymetrix: https://cen.acs.org/articles/94/i14/Thermo-Fisher-wins-contest-Affymetrix.html

[6] Tribecode review: https://thegeneticgenealogist.com/2015/08/01/a-review-of-tribecode-by-centrillion-biosciences/

[7] Funding rounds:
https://www.crunchbase.com/organization/centrillion-biosciences

[8] https://www.genomeweb.com/business-news/affymetrix-thermo-fisher-rebuff-origin-technologies-purchase-offer

Key quotes:

In reviewing the proposal, Affy said that it has concluded that Origin appears to be a “newly formed shell entity with no assets of which Affymetrix is aware, and whose sole source of funding for the proposed transaction is $1.5 billion in potential debt commitments.” Moreover, this $1.5 billion is not enough, according to Affy, to cover the proposed deal in full.

“The proposal put forth to Affymetrix by Origin Technologies, a newly created shell entity relying on a vague and insufficient financing package from a Chinese firm, is highly uncertain and speculative and does not constitute, and could not reasonably be expected to lead to, a superior proposal under the merger agreement, and Affymetrix and its board of directors could not reasonably determine otherwise,” Marc Casper, president and CEO of Thermo Fisher, said in a statement.

[9] Patents list the Centrillion address as:

Centrillion Technology Holdings Corporation
Maples Corporate Services Limited
PO Box 309, Ugland House
Grand Cayman, KY1-1104(KY
UNITED KINGDOM

Click to access Patent_18052018.pdf

[10] Funding history: https://www.whoisraisingmoney.com/centrillion-technology-holdings-corp

[11] http://www.xdata.co/story.php?a=000147383518000004&c=1473835
“Centrillion Biosciences received two commitments for its $31.59 million financing round. Investors committed to buy 89% or $28.01 million worth of equity three weeks ago. Based on the offering’s structure, the company has until June 2019 to raise an extra $3.58 million. A total of five unregistered securities offerings closed by the company raised an estimated $58.78 million.”

[12] https://www.sbir.gov/sbirsearch/detail/700133

[13] https://www.molecularvisionlab.com/mvl-vision-panel/

[14] http://www.freepatentsonline.com/WO2012134602A2.pdf (2012)

[15] https://patentimages.storage.googleapis.com/d2/b9/a5/3aea1e5c2405b8/US20120083417A1.pdf

[16] http://www.centrilliontech.com/wp-content/uploads/Centrillion_FactSheet.pdf

[17]
Selected Glassdoor reviews:

1:
Cons

Every con mentioned here is true (I advise only looking at the one star reviews for an accurate portrayal of the work environment).

Advice to Management

At a minimum, fix what is mentioned in these reviews. Management and HR need to be addressed and your employees are miserable. Stop trying to get your current employees to write reviews on this app to raise the company ratings. Instead of trying to deceive future applicants, redirect your efforts to focus on your current employees’ well being.

2:
I worked at Centrillion Biosciences full-time

Pros

Some people on the exec team are the best mentors you can have. They are encouraging and do care deeply for the employees. At times they have tried their best to shield us from the absurdity of the other half of the exec team.

Cons

There are so many cons about this company that many have already touched on in past reviews. I will point out the two below that is most concerning to me.

I worked at Centrillion for 1.5 year before leaving for a better opportunity. Since my departure, I have heard some things have gotten better where others have declined. Centrillion is a very secretive company with absurd company polices. One of the craziest company policy is clocking in and out as mention in other reviews. Everyone at the company except the interns are all on salary but it is required that you clock in and out everyday for 8 hours each day. If you are not physically in the office for 8 hours, you will be reprimanded by HR.

Starting salary is low — starting salary for research assistant can be as low as $50,000. There are ‘yearly’ raises but again it is barely enough to live on in the Bay Area or enough to keep up with the standard.

Advice to Management

Management does not listen or care as long as they can save a quick dollar.

3:
I have been working at Centrillion Biosciences full-time (More than a year)

Pros

My paycheck never bounced…
Free lunch – but who doesn’t have that?

Cons

Poor communications
Paranoid management
Managers unclear about roles
Teams not sync’d up
Operations and HR are unprofessional
Any positive reviews you see are posted by the company

Advice to Management

Move to China where you spend your time anyways

4:
Pros

– Most of the people are smart and easy to work with.
– Some of the senior management tries.
– $200 towards lunches.

Cons

– Many people are unhappy and there is an air of negativity creating a lack of motivation or enthusiasm.
– CEO micromanages everyone and pins employees against each other making it highly competitive among members of the same team.
– Secrecy is highly encouraged.
– Many are not sure why they are working on certain projects.
– There is one woman managing operations, facilities and purchasing.
– Level of education plays a major roll in deciding title, pay raise and promotions- not level of experience.
– Perks are lacking compared to other companies in the area.
– There are internal cameras pointing towards employee seating area.
– Full time employees must work a full 8 hours on a daily basis excluding any breaks taken or lunch.
– Average amount of time an employee stays with the company: 6 months -1 year.
– Management forces you to take few days of employee earned PTO during christmas break every year.
– HR creating false reviews on glassdoor to boost the overall rating!Show Less

Advice to Management

Treat your employees well if you’d like to retain them.

5:
I worked at Centrillion Biosciences full-time (More than a year)

Pros

-Some of the employees here are okay to work with
-Catered lunches but only about 80% of the month

Cons

-Retention rate of employees is discouragingly low (<1 year) and no concern is shown by management
-Very little opportunity for growth (again, management does not seem to care or prioritize their talent)
-Compensation and benefits below industry standard. The company is stingy with compensation.
-Unprofessional Operations group that was also acting as an HR group (concerns and questions are either ignored, delayed, or show lack of interest)
-The way the company is set up creates a paranoid atmosphere in the company (surveillance, clocking in, internal discussions discouraged, feels like you’re being monitored)

Advice to Management

-Invest in employee’s growth and allow them to have competitive wages to afford Silicon Valley housing
-Executive group needs to be more transparent with empolyees

6:
I worked at Centrillion Biosciences full-time (More than a year)

Pros

There are some very good people working there, and the funding appears to be adequate to make payroll. Annual raises for most employees. Free catered lunches.

Cons

Company appears to be a front for sheltering Chinese capital investors rather than an actual product or technology driven enterprise. Management is not interested in employee retention or development. Average employee duration is about one year. Essentially no path to advancement, a secretive and manipulative executive group operating without clear principles or ethics. Compensation and benefits are sharply below the local industry standard. Non-scientific sweatshop atmosphere, junky and poorly organized work areas, internal discussions are discouraged, surveillance cameras pointed inward, rank and file employees are required to clock in and out to leave the building.

Advice to Management

Stop trying to hide what you are from employees, nobody who has worked there more than 3 months believes what you say. Take some business psychology courses.

7:
I worked at Centrillion Biosciences (Less than a year)

Pros

Some good people to work with
Decent place to gain some work experience
Lunch provided by company although that is a norm in Bay Area startup

Cons

Horrible company culture

The worst HR policies. There is absolutely no concept of work life balance. Everyone is required to clock in and out. The management tells the employees that it is for “safety” but the HR keeps tabs on work hours. Workers are required to be at site for 9 hours and anything less leads to censure

Very secretive culture spread by the upper management. No one knows what the other is working on or even what the product is.

Advice to Management

Your employees are your peers not your servants. Treat them with respect.

Business

There’s not much business information available on Caerus, they’ve received ~1.2M USD in SBIR grants [1] most recently in 2014. I can currently only find one employee on LinkedIn (one of the founders, Javier Farinas), though I can find 3 other previous employees (including the co-founder, Andrea Chow). Caerus received their first grant in 2010, and I would guess they would likely need to raise soon to sustain R&D work on their platform.

Technology

Caerus originally proposed using a method they called Millikan sequencing. This was a method measuring the charge of a strand as nucleotides were incorporated. They published a proof on concept of this system in 2015 [2], however they also announced that they were abandoning the approach around this time [3].

The Millikan sequencing approach was pretty interesting. In the proof of concept they present a system where a bead covered with multiple DNA templates is suspended between two electrodes.

They apply an AC bias voltage and can see the bead moving under the changing electric field (the DNA being negatively charged). The proof on concept uses sequencing by synthesis. As bases are incorporated the charge on the bead changes and these incorporations can be detect by changes in the motion of the bead (they use a velocity measurement).

It’s pretty neat that they could use natural, unlabeled nucleotides (which could potentially lead to longer reads). But they have apparently abandoned the approach. Because they used amplified DNA, they would be subject to the same phasing errors that limit read length on other platform. Ion torrent also used unlabeled bases and I guess they figured the competitive advantage was too small. Having an optical sequencing technology that uses unlabeled bases is kind of neat however (but I guess also similar to 454s platform).

After abandoning the Millikan approach, they started working on something called “activator sequencing”. This method is covered in their 2014 patent [4].

The patent states “the methods use an enzyme to covert a product produced from a sequencing reaction into many copies of a readily detectable report molecule”. Essentially what they appear to be suggesting is single molecule sequencing [5] system where incorporation activates an enzyme which in turn creates many copies of a detectable (e.g. fluorescent) molecule.

It seems like an interesting idea, but at the single molecule it’s possible that engineering efficient enzyme activation might be problematic. It will be interesting to see how things progress!

Notes

[1] https://www.sbir.gov/sbirsearch/detail/12662

[2] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4560655/

[3] https://www.genomeweb.com/sequencing-technology/caerus-molecular-explores-activator-sequencing-long-single-molecule-reads-low

“For amplified DNA, they did show proof-of-principle of sequencing, which they published last month in Analytical Biochemistry. However, it did not make sense to develop that approach commercially. “It was pretty evident that we were not being able to catch up to the rest of the field using that method,” Farinas recalled, noting that Ion Torrent came out with the PGM around the same time, which offered many of the same benefits, including label-free sequencing, that Millikan sequencing promised. “Basically, they had a commercial product and we had some preliminary data,” he said. ”

[4] http://www.freepatentsonline.com/20160068902.pdf

All Patents:

Click to access 20120220486.pdf

http://www.freepatentsonline.com/y2016/0068902.html
http://www.freepatentsonline.com/y2010/0112588.html
http://www.freepatentsonline.com/y2011/0059864.html

[5] The patent also suggests they the scheme could be used with colonies (clusters).

It might seem odd for Bionano to be on a list of DNA sequencing companies given that Bionano have not publicly discussed the developing a DNA sequencing platform. Bionano are instead focusing on genome mapping. Mapping technologies give long range information on the structure of the genome (megabases, at >100bp resolution). This allows you to detect large scale structural variation in a genome (such large scale changes are associated with cancer for example). In contrast to this, most current DNA sequencing technologies focus on accurate short range information, providing single base resolution reads of about 100bp.

However mapping and sequencing exist in a continuum. As sequencing read lengths increase (into the mega base range with some approaches) they can more easily address issues of structural variation. Similarly, as the resolution of a mapping technology decreases (ultimately to single base resolution) it turns into a long read sequencing technology.

Because of this technological continuum, I’ve included mapping companies that have their own unique sensing system. And in this post I discuss Bionano Genomics.

Business

Bionano genomics (originally Bionanomatrix) was founded in 2003, to date they have raise 132.1M USD (according to crunchbase) from a number of investors [1], they also received about 3M USD in SBIR grants [3]. Their first commercial instrument launched in 2012. They now appear to be filing for an IPO seeking to raise 34.5M USD [2]. Genomeweb also reported that Bionano received 1.7MUSD in revenue Q1 2018 (net loss 3.8M). On March 31st they had 7.6MUSD in cash. Meaning without further financing they can last a couple for quarters. In 2017 total revenue was 9.5MUSD. So the 2017 average revenue per quarter was a little higher than Q1 2018s. They currently have 65 employees (28 in research).

Technology

Bionano’s devices were at least partially developed from work undertaken at Lund University and Princeton. Patents refer to a 2004 paper co-authored by Bionano Genomics founder Han Cao [5]. The paper shows 100nm nanochannels:

In this paper double stranded DNA is driven through the nanochannel under a bias voltage. There’s a nice image of some stretched DNA:

Here is DNA is labelled with TOTO. This is an intercalating dye which labels the entire strand. The Bionano chip seems to be a development of this, and there’s a nice schematic on the Bionano website:

As you can see the chip is somewhat more advanced than that presented in the 2004 paper. In particular pillars have been added to untangle DNA. Once the double stranded DNA is linearized in the channels it can be imaged.

But just imaging stained DNA isn’t very useful. In order to create a structural map of the fragments, we need some kind of specific labelled to create a “map” of the fragment. Most typically this is performed using a nicking endonuclease. These enzymes recognize particular target DNA sequences and create small single stranded “nicks” in the DNA:

 

 

Fluorescent probes can then be attached to the nicked locations to allow their locations to be imaged. Using this information you can obtain a map of the DNA strand, telling you the relative locations of the nicking enzyme recognition sites. You can then compare these molecule maps against known references. By looking for large scale rearrangements, it’s possible to determine where structural variation occurs.

The system resolution is likely limited by several factors, and their site states that the Irys system resolution is 1.5 kbp or ~500nm, this would make sense as the optical resolution of a diffraction limited system.

In addition to there currently released platform Bionano do have one patent where they incorporate sequencing into their platform [4], but given this is from 2009 it seems likely that they have not pursued this approach currently. Some more digging might reveal other interesting approaches in their patents.

Notes

[1] According to Crunchbase, investors include: Domain Associates , Novartis Venture Fund , Legend Capital , Battelle Ventures , Gund Investment, LLC , KT Venture Group , Innovation Valley Partners , 21Ventures and Ben Franklin Technology Partners of Southeastern Pennsylvania.

[2] https://www.genomeweb.com/business-news/bionano-genomics-seeks-raise-345m-ipo

“Bionano also revealed that it had $1.7 million in revenue in the first quarter of 2018 and a net loss of $3.8 million, or $1.16 per share. Its R&D expenses for the quarter were $2.4 million and its SG&A expenses were $2.9 million. As of March 31, the company had $7.6 million in cash and cash equivalents.

In 2017, the company had $9.5 million revenues and a net loss of 23.4 million, or $7.66 per share. Its R&D expenses were $12 million in 2017, and its SG&A expenses totaled $14.1 million.

As of December 2017, Bionano had raised approximately $129.3 million through sales of its preferred stock. In addition, in 2016, it entered into a secured term loan facility with Western Alliance Bank, under which it has borrowed $7 million. The loan facility requires it to raise $5 million from the sale of equity securities by Aug. 3 of this year.

Bionano said that as of March 31, it had 65 employees, including 28 in sales, sales support, and marketing, 28 in research and development, four in manufacturing and operations, and five in general and administrative positions. Of its employees, 57 are located in the US and eight elsewhere.”

[3] https://www.sbir.gov/sbc/bionanomatrix-inc

[4] https://patents.google.com/patent/CA2744064A1

[5] http://www.pnas.org/content/pnas/101/30/10979.full.pdf