It’s getting late and I’m on my 5th disc download of RedHat RHEL-5.7 i86_64. Please stop me now if this is not the right version for the MacBook Pro i7 processor because I am running out of blank discs. I didn’t want to go with the beta versions because I have a lot of applications to test and beta would only complicate things.
The screendoor is open. For the first time in more than a month, the air conditioner isn’t droning and drowning out the sound of the world outside. Tonight, I hear the rain tapping low and high notes on my weatherbeaten deck. While I wait for the binaries to download, I’m sipping a bit of white wine and listening to the watery music.
Break open your best boxed Merlot and join me. Tonight, Eva Cassidy sings, life goes on and the loveliest sounds in the world can not be bought for any price.
Today is my “Release from Wage Slavery!” day. There’s a bottle of bubbly in the fridge left over from New Year’s. I’ve debated whether to drink it now or wait until I land another job. The most ironic thing about being laid off for me was that I got so busy after it happened. In fact, I didn’t get it all done. There’s a structure that needed a few more steps of refinement and at least a day of tweaking. Nah-gah-happen. Then there’s that threonine that’s taunting me. Arggghhh, it kills me to leave it all hanging like this just when I thought I was onto something.
Last night, I followed a link from Derek Lowe’s blog, In the Pipeline, to a cathartic post from Anthony Nicholls who runs OpenEye. Here comes some geek speak: I’m a big fan of OpenEye. They make an application called ROCS that allows for shape based searching of chemical structures in 3D databases. I’d put ROCS right up there with SMILES in terms of computational chemistry innovations. Once you’ve gotten the hang of it, it’s hard to see how you can live without it. It’s versatile like an iPad. There’s no one way to use it. The variations are endless. This was brought home to me yesterday as I was closing out my notebooks (yes, I’m one of the few modelers who even keep the damn things). I was indexing the entries and couldn’t help but notice how many times I had used ROCS, OMEGA and most recently, BROOD. Excellent tools. I’m really going to miss using them. Next to some more bloated software suites, OpenEye is cheap. A great bang for the buck. But I can’t afford even a single user license on a severance package budget. (There might be a business opportunity for Anthony in that for the newly independent modelers who need to keep their skills fresh until they land their next gig. hint, hint) Ok, enough of that geeky stuff.
Anyway, back to Anthony’s rant, What’s Really Killing Pharma, I think Nicholls nails it- almost. There is something he left out that puts the whole demise of pharma into perspective. (For those readers out there who are lifelong enemies of pharma for ideological reasons, please don’t get in a snooty snit over it or you will be missing the point of Nicholls’ post. This problem goes beyond the misperceptions about “me too” drugs and outrageous pricing.) The whole post is worth a read but some parts stand out for special attention. It’s almost as if Nicholls has been spying on us (it’s funny how he can’t resist adding references even in a blog rant):
But what has actually happened? First, a lot of senior scientific talent has moved on—some through retirement, others preferring to work at smaller companies with less of the “world-class management” big pharma can provide. But what has shocked and distressed me is the number of people in their fifties who have been let go. These are the people who actually have a working knowledge of the fifty years of pharma Witty mentioned, people who have done their “10,000 hours” [2] refining unique and irreplaceable skill sets, people who can pass these skills on to others. If you accept that making drugs is more art than process, then these are the last people you let go.
But even this travesty pales next to my next point: the danger of management fads. Because your modern big pharma CEO knows next to nothing about science, I have to assume they think they are adding value by imposing management schemes they do know about. Let’s consider one such disaster of a fad: lean thinking and six sigma. Originally developed at Motorola by Bill Smith but based on earlier concepts from Genichi Taguchi and others, the concept is simple enough: apply statistical modeling to an industrial process so that one can gradually improve that process. Actually, this principle is not dissimilar to my decrial of the lack of statistics in molecular modeling—if you don’t know where you, are you never know if you have improved. The problem, though, is the process being modeled here—drug discovery—doesn’t lend itself to this method. As any senior medicinal chemist or molecular modeler would be happy to explain to management, an embarrassingly large fraction of drug discovery involves serendipity—while you’re looking for one thing, you find another. And serendipity is, of course, the complete antithesis of a Taguchi robust process [3] where variance, i.e. a standard deviation, can well defined- we work in the domain of the unexpected, the domain of the “Black Swan” [4]. Now that the method has been applied and failed, it seems ridiculous to have ever thought it might have succeeded. But not only was it applied with great vigor, it often came to be seen as a much more secure employment path than the vagaries of drug discovery. Not a little talent was wasted on these meaningless exercises and not a few careers lost to management bullshit.
Another good one: empowering IT departments to make scientists use the same infrastructure as the guy at the front desk. Rather than see that scientists often have different computing needs than other parts of the business, IT demands obeisance to the corporate norm. In doing so, they hinder the kind of innovation (e.g., Linux, GPU solutions) that used to regularly occur because scientists are quite computer literate, thank you. Instead, IT departments make it impossible for competent people to manage their own resources. They create obstacles instead of removing them. Machine was made for Man, not Man for the Machine.
More fads? How about metrics and cross-charges? In this Through-the-Looking-Glass world, scientists have to account for everything they do, with the cost of each and every action weighed and accounted for. Work done by other groups is counted as “services” that have to be expensed. In other words, upper pharma management—convinced, perhaps, that scientists don’t know the real-world cost of operations—are ruining the one indisputable advantage of a big company: the fact that you can just walk down the corridor and get the another person’s expertise to help solve a problem. They are building walls that turn a large company into a thousand little independent entities with all the problems of communication and lack of shared vision that implies. That’s the newest, most amazingly dumb-headed, most disastrous strategy. Now tell me, does this empower your researchers? Does this “remove obstacles”?
That was orgasmic. I need a cigarette.
I have *been* there, especially with respect to fighting the corporate IT department. How much money and wasted time has been spent hiring contractors, who don’t have the security clearance to get the sys admin password, to manage our linux workstations and clusters to corporate IT standards. I can remember when I first started my current gig when we were transitioning from IRIX to RedHat and I had to deal with some pretty clueless borgs in IT purchasing who weren’t going to let me order the new workstations for our group because the only computers we had a contract to purchase had to have Windows on them. We finally got the workstations and had to have Windows removed, which violated a warranty. It was ridiculous. So much time and email spent trying to get the IT department to understand why we needed a different operating system. If pharma wants to save money, they will get rid of the bulk of their useless, paper pushing, “we only do one thing”, IT bureaucrats and give in silico sciences the things they desperately need- their own sys admins and private networks separated from the rest of the accountants and salespeople. And for God’s sakes, get rid of IE6. No one uses it anymore. Please don’t tell me you’re only interested in security when you shut down my Chrome browser and then force me to do BLAST searches on IE6. No amount of fencing the campus and card swipe enabled portals are going to keep determined bad guys out of your databases if you can’t keep up with the latest security features of your old Windows operating system.
And don’t even get me started with the charge backs to other departments scheme. It’s stupid beyond belief for research. And here’s why: research happens by trial and error. Sometimes many trials and errors. (Actually, I’m encouraged that statistics and “design of experiments” is starting to catch on in protein engineering and expression but even using “design of experiments” is going to take trial and error. There’s no getting around it.) Yes, it can be expensive but if you have to limit your trials because a.) your budget has been unrealistically reduced or constrained and b.) negotiating every transaction takes away from your time at the bench, then you have effectively strangled the baby before it has a chance to grow. I’ve already seen this in action. Project teams want and need your expertise but the company is forcing them to pay for it out of a vastly reduced budget and they have to weigh every choice very carefully. What happens next? The project team is either forced to forgo this vital knowledge or pay an outside company to do it. But an outside company is constrained by science as well. THEY need to go through a process of trial and error too in order to deliver what the project team has ordered. And that means the project team will get a limited number of attempts to get the information they need because otherwise, the outside company is going to start charging them for everything that goes out of the scope of the negotiated contract. AND, here’s the real kicker, the outside company can sell that protocol to some other company. If your project team wants to keep the information exclusively, there is an additional charge. How is this cheaper than going down the hallway to ask if the hard working little department (that is getting laid off) if they can do it for you?
BUT, there is one aspect of this problem that Nicholls has missed (or maybe it’s in his next missive on the subject). The problems that Nicholls has illustrated are not unique to pharma. Maybe they have a disproportionate effect on pharma because some parts of the process are more sensitive to external pressures. But in general, the effects of mismanagement by the MBA Bonus Culture can be seen in every industry from banking and finance to automobiles and chip manufacturers to television and cable media companies.
Here’s my theory about how it all went south since the 90’s: It’s the 401K. Yep. It puts the whole dismal phenomenon in the proper framework. But how could something that appears to be so innocuous bring the country’s scientific and industrial framework to a screeching halt, you ask? Think about it. MBAs and executives are hired by corporations to “increase shareholder value”. Before the rise of the 401K, that meant building a better mousetrap and finding a better aspirin. Now, the only thing that matters is pleasing the finance guys who rate your stock from buy to hold to sell. And those same finance guys have a stake in the outcome. They set their own terms of compensation as well as service large institutional investors and mutual funds that we get to choose from in our 401K plans. Corporations are now driven to serve the finance industry and the shareholders, which are us. They’re not in it for the new products anymore. And once you start hiring executives whose goal it is to optimize the ROI, well, it doesn’t matter where the money comes from anymore. It’s just money, numbers on a spreadsheet. It doesn’t matter if those numbers represent people with families and caloric intake requirements and 10,000 hours of expertise. It doesn’t matter if there are fewer products to sell. Once the patent is dead and the money is gone, the investors will follow the money to the next hot thing. There’s nothing mysterious about this. It’s not personal. It’s just the reward system we have set in place since pensions became too old-fashioned for those up and coming 30 somethings in the 90’s (um, that would be us). We’re like rats pushing a pellet bar for another shot of cocaine. We can’t help ourselves and will keep doing it until we die from the pure pleasure of getting that next teeny bump to our retirement accounts.
Now, is the 401K the only reason why pharma and other industries in the world are starting to topple and fall like Ozymandius? No, of course not. There are other forces at work here, like the FDA that really needs to get its act together and out of control lawsuits that are affecting everything from pulling pretty safe drugs off the market but leaving stuff like Tylenol available for anyone to buy off the supermarket shelf to potentially destroy a liver or two to making it impossible for school age kids to have a normal life of scabby bruised knees after a vigorous afternoon at the playground. We lefties are partially to blame for the overly litigious nature of American culture where everyone has an opportunity to plunge their hands into the deep pockets of someone else’s bank account or insurance policy for the slightest of injuries. How many of us realize that the money guys just pass the costs onto the consumer and keep on partying?
But I have run out of time this morning. I have to go to work, for the last time in what may be a very long time, and finish transferring my files. Then, I will drop my keys, SecureID, credit cards, dosimeter and employee badge into an envelope and say good-bye to the best job I ever had.
The Confluence 