With the "How Will Bioinformatics Scale?" panel discussion coming up, now is a good time to reflect on the last year of DNA sequencing and bioinformatics industry gyrations. Besides, we're already 1/12 through the year, so one had better reflect quickly before last year and this year blend together.
(Can I pretend that such time-blurring is in fact an advanced subconscious shortcut to update the mental models without burdening the model skeleton with disjointed, soon to be deprecated facts, rather than a possible case-history indicator of APOE status?)
Anyway, my observations over the last year are necessarily colored by a distinct commercial perspective. Caveat reader.
Sequencers for the clinic
I should perhaps start and finish with Illumina ($ILMN), but I won't. (Or did I just?) Rather, QIAGEN ($QGEN) is the intriguing actor this year. Despite little visible progress on bringing the GeneReader to market (though there were some nice mechanical shells at the AMP meeting in Phoenix), they demonstrated their sincerity by acquiring the leading pathway analysis and desktop sequence analysis bioinformatics companies: INGENUITY and CLC bio. INGENUITY, especially, speaks to $QGEN's laser focus on powering routine clinical molecular diagnostics. They didn't take long to follow up with an early-access program for the clinical version of their Variant Analysis web tool, with a killer roster of clinical/scientific advisors. My impression: if anyone is going to be a serious contender for arming the world's molecular pathology laboratories with NGS-based diagnostics capability, it's $QGEN. Even without the GeneReader.
Bioinformatics in the Cloud
While $ILMN made an interesting counter-acquisition in NextBio, I found the steady progress of BaseSpace's capabilities more important. While DNAnexus was getting v2.0 of their platform off the ground, StationX was prepping and launching GenePool, and other ab initio or grant-derived players (Seven Bridges Genomics, Maverix Biosciences, Arvados, GeneStack) were escalating the cloud-based informatics war, BaseSpace was relatively quietly evolving, sheltered safely under their 1-button ("check here to upload to BaseSpace") ease-of-use for data capture. It's usability hooks like that we need to watch.
The industry has very quickly fronted a diverse cast of contenders offering similar levels of functionality on virtually identical infrastructures, raising the bar quite high. Not bad, considering only two years ago, "the cloud" was not supposed to be fit for genomics.
Not to gloss over the piles of interesting papers published over the course of the year, but, some of the more interesting journalism came from big research: ENCODE, and Dan Graur's colorful deconstruction of it. Or, how about TCGA, and the realization that no one can easily use the data (see NCI's $20M call for interest on building ... wait for it ... a cloud based informatics platform).
Other sequencing tech goes quiet
Because the sequencing instruments are already so good, I think we all sort of stopped agonizing over it. After resetting shareholder expectations, Pacific Biosciences ($PACB) continued to iterate and tailor their areas of expertise. Oxford Nanopore continued not to offer data (though brilliant marketing, again, at ASHG, by announcing a signup sheet for a beta program; one step shy of just putting the whole MinION on IndieGogo). What about Life Technologies ($LIFE)? Caught up in acquisition news and with its loss of marketing momentum on the Ion Torrent side, its position as second-place lead in the sequencing market is starting to feel more like the first loser position.
Software investment ramps up
As venture investors start doubling down on their bioinformatics portfolios (rounds for Spiral Genetics and Bina Technologies and SVBio and Knome and Syapse and Personalis and Curoverse and more), everyone is shifting their rhetoric to focus on clinical informatics.
A question in my mind is whether clinical bioinformatics will turn out highly fragmented and chaotic like research bioinformatics, or whether it will coalesce around a couple of preferred vendors. (Full disclosure: my current startup, DxEngine, is developing software solutions to help pathology labs produce actionable clinical reports.) Perhaps the only winners will be Amazon Web Services ($AMZN) or Google Compute Engine ($GOOG), selling the computing infrastructure pickaxes to the bio data miners.
The FDA has obligingly offered brilliant political doublespeak with respect to LDTs, connected devices, EHRs, and the like: they seek to reduce regulatory uncertainty by claiming carte blanche jurisdiction, issuing guidelines, and then draping it all with "enforcement discretion." Thumbs-up for guidelines, but pardon me for not feeling certainty when enforcement policies are riddled with discretion. Nonetheless, double thumbs-up for an FDA-cleared MiSeqDx.
2014 off to the races
Which brings us into the beginning of 2014, where we can already see a continuation of double-down investments, $ILMN making serious instrument waves, and re-excitement for delivery by Oxford Nanopore and $QGEN. And, clinically, we're all watching to see how reimbursement in general and CMS in particular responds to the growing body of diagnostic tests and the subset with clinical utility.
2013 was a fun year, and if January is an indictor of 2014, I might need to do a mid-year update just to keep it all straight in my head.