GenapSys: Business Models for the Genome
I came across a case study on GenapSys on the Harvard Business Review website. I thought it would be fun to review it. Particularly given that liquidation of the company seems to be going ahead.
So I went ahead and paid the $9. The case study starts by discussing the sequencing market as of 2011, and why they think semiconductor sequencers will be disruptive:
Esfandyarpour noted that most sequencers used optical systems: DNA segments were labeled with dyes, and cameras visually examined it…. The complex nature of optical systems made such devices expensive, as well as labor- and time-intensive to operate.
For some reason it’s become received wisdom that optical = complicated. This isn’t really the case. Complex systems are complex. Simple ones simple.
To take one example. Both the iSeq and Ion Torrent platforms are disposable chip based systems. Both use sequencing-by-synthesis sequencing approaches. The iSeq however uses an optical readout.
The iSeq is a lower cost platform than any of Ion Torrents offerings, and I would say is a significantly less complicated approach… the MiSeq also uses consumer grade CMOS sensors, likely making both the platform and the consumables cheaper than Ion Torrent. Optics is only one factor that contributes to the overall complexity and cost of a platform.
Again in the preamble, the case study notes how different sequences are suited to different applications:
A machine’s efficacy for a given application was influenced by how DNA was prepared for and read by a sequencer. For example, one group of researchers found that the Ion Torrent platform was less adept at sequencing genomes made up of adenine and thymine (AT) base pairs, which could be an important factor for those working in specific fields, such as the study of malaria, as the researchers noted that over 80% of the malaria parasite’s genome was composed of AT base pairs. After testing a number of NGS platforms, the researchers concluded that, “The decision on whether to purchase one or the other [sequencer] will hinge on available resources, existing infrastructure and personal experience, available finances, and the type of applications being considered.”
This kind of hand waving is pretty common. It’s easy to hedge your bets and say “it depends”. The reality is that since 2011, it’s been hard to justify the use of Ion Torrent systems for technical or cost reasons (business and other factors may come into play).
The case study then goes on to discuss 3 possible models for GenapSys:
Traditional Sales Strategy - $40000 instrument, $500 to $1000 cartridges. $10000 servicing fee. Sell to research institutions.
Razor-and-Razor-Blade - $5000 to $7000 instrument, $200 cartridges at 50% margin.
Data Analytics - $3000 instrument or $2000/y lease. $50 to $100 cartridges. Customers give GenapSys copy of all data generated. “We could become the Google of genomics.” Easfandyarpour said. GenapSys would primarily derive its revenues from test royalties and from licensing or charging a subscription fee for access to its database.
Here’s what Esfandyarpour had to say:
Esfadyapour explained his views on each model: “The traditional model is a financially safe route. It’s a proven strategy and there’s a good chance we’ll be acquired, but I don’t see this as the way to change the future of medicine. With the razor-and-razor blade model, GenapSys can bring an affordable sequencing technology to new markets. However, it will ultimately be the company that controls the information that will succeed and change how healthcare is delivered.”
Ultimately GenapSys launched a $10000 instrument with $300 cartridges. I didn’t hear anything about a “data analytics” strategy. But then again, I’ve not really heard anything from any GenapSys customers… so they ended up somewhere between “traditional sales” and “Razor-and-Razor-Blade”.
While Esfandyarpour felt the “traditional sales strategy” would have been a safe route. I’m not so sure, at the price point suggested there would have been no strong advantage to the platform over an Ion Torrent instrument (or really, a MiSeq). And the fact that the company didn’t even end in acquisition suggests that no other players were particularly interested in the technology.
The “Razor-and-Razor-Blade” model clearly didn’t work out. It doesn’t even seem to work that well for a $1000 MinION with $100 to $900 runs (which doesn’t seem to generate significant revenue for ONT). I suspect to make this work you need to get run costs much lower than this and expend significant effort developing new applications around the sequencing platform to drive sales.
The “Data Analytics” model hasn’t really been tried. I suspect it’s very hard to find a sufficiently large market to engage with this model, both on the user side and in terms of users willing to pay for access to your database.
If you can push below a $100 COGS (closer to $10) and sell runs for $20, you might be at a price point where you can start to compete with qPCR and open up new applications. But this requires significant market development, and is far from an easy sell.
I do hope we see more cheap sequencers in the future but at present few players seem very interested in even developing MiSeq class instruments… I explore this, and other issues around GenapSys after the paywall break below.