Monday, September 29, 2008

The paper from Hell

So, the hike was wonderful. Beautiful Redwood forests, happy lizards, dragonflies, blue skies, views of tree-covered mountains sweeping down to the misty sea and so forth, good company and nerdy jokes and wonderfulness. On the way back was a cloud that looked like a purple bruise on the blue sky. But that's not what I've been thinking about today. Today I have been thinking about (or trying to think about) a paper assigned for tomorrow that simply would not unravel its science to me.

In fact, it was only when I realized that the paper wasn't one research study but four independent experiments that I could understand what on earth was going on. At which point, since none of the experiments were particularly compelling, I was left with an even more sour taste in my mouth about the whole experience. Perhaps they assigned this paper simply to frustrate us; to see who could read a paper and who couldn't, who could determine the weaknesses and analyse and generally understand the steps of a paper. And to see who wasn't afraid to look at a published work and say "That paper was crap." Because this one was particularly hard to understand: no organization, no background, and no continuity from one part to the next.

As best as I can reproduce it, for your reading enjoyment, here is a summary.
First off, most people think that this one protein (A) interacts with the cellular scaffolding proteins to make cells asymmetric (btw, cells are asymmetric. nifty.) Except, if you prevent scaffolds from forming, you still get asymmetry, so this can't be it.

If you look at a cells that mess up their asymmetry -- don't bud/divide correctly -- you find that a lot of them have problems with another protein, B. (This protein is totally unrelated to protein A.) If you cut out B, along with some other proteins that let the cell know which way is up, cells become symmetric. If you add back B, they lose their symmetry. If you add back various mutated forms of B, effective and deficient in various functions, you can see that certain things, like binding to the cellular scaffolds, are necessary for cell survival and asymmetry, and other things, like binding to itself, are not necessary for anything in particular as far as we can tell.

Protein B might interact with protein C. Protein C is thought to serve as a +/- marker, like an on-off switch. But it might not work quite that way. If you force it to be a switch, you get symmetry, so it's most likely not a switch.

By the way, if you add lots and lots of the switch-version of Protein C, you can in fact get asymmetry back. But those cells aren't happy, so they don't count.
Honestly, I shit you not. Now put that over 13 pages of techno-babble with appropriate (or inappropriate, you know, to spice things up) figures, and you get the paper I had to slog through today. Delightful, right?

The one thing I take from this paper is that as a biologist, I need to learn how to design a study, in a way that is not just a collection of various and sundry experiments, because otherwise I will end up writing a paper like that one.

And that would be very, very saddening.

Also that stream-of-consciousness, while very effective as done by James Joyce, is not an appropriate literary device for a scientific paper. But I already knew that.

1 comment:

Duff said...

Papers without view toward a specific problem are terrible reads.