Michael Davidson on knowing how things work

“In competitive biology, you don't necessarily need to be a mechanic but you need to be able to operate the machine...If you don't know how it works you'll get creamed in the race."

- Michael Davidson, as quoted in Nature

In this week’s Nature, an editorial sounds a cautionary note about scientists who use complicated machinery but don’t really understand how it works. Whereas physics has a tradition of building its own apparati and rejoicing in its complexity, biologists often don’t know the first thing about how their ‘black boxes’ (many of which were designed by physicists) actually function, and what their possible artefacts might be.

Further on in the issue, journalist Helen Pearson highlights fluorescent microscopy as a great example of the dangers involved in dabbling in a technique without proper training. In a culture when authors, referees and journals are too often “impressed by pretty images”, and everyone is in a hurry, accurate analysis can take a back seat. The technique is fraught with pitfalls, especially FRET (fluorescence resonance energy transfer), but even the humble bread-and-butter technique of co-localization (where red-tagged and green-tagged proteins are inspected for the telltale overlap color of yellow) is often done improperly – perhaps up to half of the time, one expert estimated. But University of Florida microscopist Michael Davidson (quoted above) thinks there is an "infinite number" of even simple variables that can also cause trouble, and another expert consulted in Pearson’s article thought that a substantial number of published results (5-10%) based on fluorescent microscopy might actually be wrong.

So what’s the solution for biologists out there who want to do it right? Take a serious course, consult handbooks, collaborate with experts or outsource the work. And how do journals police poor microscopy? Force authors to give more details on the conditions they have used to create images, and employ at least one referee who knows microscopy backwards and forwards to spot shoddy work.

Both Nature articles are available online with a subscription.