Posted By Caulimovirus on January 22, 2010
.: The other day the thermal cycler in my lab did something unusual: it finished its run in just over 10 minutes. This is unusual because runs on thermal cyclers generally last 1 to 2 hours. Normally after I press “Start” I’m out the door and headed for lunch, but this day I stayed for a few minutes to rinse a flask and happened to glance at the display. Had I not stayed the extra moment, I would not have noticed that this particular run was abridged and probably would have proceeded as usual until eventually confronted with a baffling batch of results.
.: The anomaly had me curious. I checked the program and everything was entered correctly; it was the same program I had used half a dozen times without issue. I ran it again with dummy samples, and it worked perfectly. So why did it screw up that one time? Or, phrasing the question in a more philosophically defensible manner, how could it screw up that one time?.: Here I must admit ignorance. I’ve no idea how thermal cyclers work. Consequently, I have no idea how thermal cyclers break (beyond the obvious “the lid snapped right off” scenarios). I don’t know how they’re built. I don’t know the software used to program them. I don’t understand the engineering principles on which they operate. Though there is a small inclination within me to figure this stuff out, I simply don’t have the time to do so. I have to chalk it up as a mystery and move on. It’s an utterly banal mystery, to be sure, but I’ll still likely never figure it out.
.: Going forward, this isn’t likely going to be a problem. If, later, I get products from another reaction, I know the thing worked. There’s only one way to amplify lots and lots of DNA and that’s with the aid of a functioning thermal cycler. The problem is going backwards — when things don’t work. Say I don’t get products from the reaction; is it because the thermal cycler acted up again? Maybe. Or it could be a bad stock of enzymes that were left out to thaw overnight. Or it could be an mis-calibrated spectrophotometer telling me I have a lot of DNA when I don’t. Or it could be that nature fundamentally disagrees with the aims of my research — how would I know?
.: The answer, of course, is “controls”. But controls are expensive. Sure, I’m blessed with less uncertainty with every experiment if I use controls, but omitting them means I have more reagents to do even more experiments! What’s more, the amount of time it takes me to set up five reactions with controls could be spent setting up ten reactions without. I’d get a lot more done a lot more quickly . . . until I hit that wall of baffling results. Then I’d have to go backwards and try to figure out what went “wrong”, only to be faced with uncertainty at every step all the way to the beginning. Such is the tension produced by limited resources and everyday mysteries.
.: Of course, this all won’t be quite as severe a problem in the future when everything is free and plentiful and there are no time constraints on any human activity — all thanks to science. Yeah.