All opinions are my own and do not necessarily reflect those of Novo Nordisk

The recent news about the United States Government monitoring a great deal of both general and specific electronic data has had one beneficial outcome (or at least, one I feel is beneficial):  it has made more people aware of what can actually be done with data, and also that we’re leaving massive amounts of personal data out there that can be traced to the behavior of individuals or organizations.  A few months ago, the Seattle Times published an article describing the explosion of big data and how that can be leveraged in so many ways.

This led me to speculate, in a very out-there kind of way, about what kinds of data Biopharma companies produce and whether there’s any hidden value in that.  Now, certainly companies are very careful about communicating information to the outside world.  Contracts with collaborators routinely contain embargo clauses, and presentations and posters are carefully vetted by legal and communications departments.  So companies would appear to be covered there.  But what kinds of data are out there that might be available, maybe not freely, but in potentia, to an interested audience?

Let me digress for a moment about mergers.  Biopharma over the last few years has seen a flurry of merger and acquisition activity.  The big pharma deals, like Pfizer/Wyeth, and Merck/Schering-Plough, have gotten big press, but there has also been a lot of consolidation among reagent suppliers.  To take one example, I’ve shamelessly taken Life Technologies’ merger history off of Wikipedia and condensed it into this table (after the jump):

Whew!  And you thought Pfizer was bad.

Given all this, there are two conclusions.  First, Life-Tech/Thermo Fisher has representation into many, if not most, domains and phases of Biopharma research.  And second, given this, this company and others like it are able to collect a huge amount of information about their customers based on what, when, where, and how much of various reagents they’re buying.

Now, all agreements between vendors and customers have terms and conditions, and what I’m about to speculate about is only valid if Ts & Cs are structured in a specific way such as to allow vendors to analyze metadata about purchasing patterns.  But I would suspect that vendors are allowed to use this data since they want to market and improve their offerings.

This data can be used the same way Walmart used consumer data to realize strawberry poptarts (and beer) should be shipped preemptively to any region that was about to be hit by a hurricane or other imminent natural disaster.  This alone is a great advantage.  Reagent companies may be able to take data about purchases from one biopharma company and use that to forecast what the overall industry might be doing in the next few years.  Using that data and analysis as a bellwether, reagent companies can ramp up offerings in that modality.  Thus, even though there may be corporate strategies that are secret, the reagent companies may be able to figure enough of them out to jump the gun and get a head start on creating the reagents that will be in high demand tomorrow.

Risky?  Sure, but as these reagent companies have become bigger and more diversified, the amount and kinds of data they can collect and cross-reference grow hugely and, I would suggest, de-risks this kind of forward looking strategy.

And what else could biopharma companies do?  Another thing could be using the information to inform investments.  If you know a given company is suddenly buying a whole lot of neurobiology reagents when they previously had not bought any before, that might suggest a push into a new indication, and thoughtful investors could parlay that into educated bets on what a biopharma company might be planning.  Such as an acquisition or an announcement, either of which could have a great impact on stock prices and reagent demands.  I won’t go further into the implications for intelligence gathering for specific drug targets, but that’s another possible avenue for a reagent company to explore since, after all, any oligo or cDNA sequence that’s ordered can be readily compared to the reference genome, and any antibody that’s bought has a known binding partner.

So is there anything to prevent reagent companies from pulling an NSA on their clients?  Well, one consideration might be good relations.  I’m sure Biopharma companies would prefer their data not be mined and, as I mentioned, there may even be provisions in Ts & Cs documents to prevent this kind of mining.  Although it also seems like proving data mining by reagent companies would be hard to prove without a lengthy court battle which no one would probably want.  Also, the consolidation of reagent companies has created a kind of quasi-monopoly where there really aren’t that many options for a lot of specific reagent types.  Scientists are often quite wedded to their favorite reagents, and shifting them to another brand is not a trivial task.  That makes cutting a reagent company off problematic.

Another thing standing in the way, though, is just feasibility.  This is a thought exercise for me, but so far I haven’t heard of any company actually doing this (although if a company is using the data to inform investments, we probably wouldn’t since it sounds a bit like insider trading).  A recent article in the New York Review of Books about the government’s surveillance efforts (HT to @ldtimmerman) has a telling quote in reference to the shutdown of a bulk email and internet data collection program in 2011: “according to a joint statement issued on July 2 by senators Ron Wyden and Mark Udall, the real reason the program was shut down was that the NSA was “unable” to prove the usefulness of the operation. ”  This suggests that all that metadata wasn’t actually that useful in finding terrorists and threats.  Same thing might apply to the data these reagent companies hold.

We’ll see if anything comes of the purchasing data of Biopharma, but at the least I think it’s something everyone on both sides of the transaction should be thinking about, because data is not what it used to be.  It’s a lot more.