Please visit our new site!

Lab Rats

The untold story

What doesn't make the cut in scientific papers

Matthew Hall 21 October 2006

Type-cast: scientific protocols are reinvented in the writing

Given that protocols aren't as useful as one is taught in History and Philosophy of Science courses, why can't they be more personal?

I've been thinking about the description of how experiments are done in science, which scientists call protocols. They’re just so dry. For starters, no humour is allowed in scientific papers. I'm currently writing a paper that includes the following statement (which every collaborator will insist is removed before submission):

Dyson and co-workers considered the postulated role of glutathione-S-transferase (GST) in resistance to platinum-based therapy and conjugated the clinically trialled GST inhibitor, ethacrynic acid, axially to create a Pt(IV) complex coined ethacraplatin (4), whose name appears to belie its abilities.

And the joke is weak. But that's as good as it gets.

What’s worse, although the idea is that in a published research paper, each experiment should be described such that anyone else in the world can read the protocol, repeat the experiment and get the same results, this isn't what actually happens in papers. First, science doesn’t always work that way. And second, sometimes scientists will leave out a crucial step in their secret recipe especially to hinder their competitors. So if you want to build on someone else's work, you need to work out how they really did it before you can move on.

But given that protocols aren't as useful as one is taught in History and Philosophy of Science courses, why can't they be more personal? Almost like a science diary? To illustrate this, let’s have a look at one of my protocols as it was actually published in 2004, then compare it to how it really happened. First the industry-standard spiel:

The lipophilicity parameter, log Poct, of the eight complexes was determined by measuring in triplicate the partition coefficients of the 1-octanol/water system based on the method reported by Robillard et al. [27]. The solutions were made up to 500 ml (100 lM) in 0.15 M KCl (7.5 : 10 2 mol, 5.59 g) in order to minimise aquation of the complexes. Equal volumes (2 ml) of the aqueous solution and 1-octanol were shaken together mechanically for 24 h. Aliquots were taken from the upper octanol layer and the lower aqueous layer. The water samples were diluted 100-fold with rigorous shaking. Platinum content was determined by graphite furnace atomic absorption spectroscopy (GFAAS). The partition coefficients were expressed as the log10 of the ratio of the platinum concentration in the 1-octanol layer to the platinum concentration in the aqueous layer according to Eq. (1).

But truthfully, that particular experiment really ran more like this:

We decided we needed to work out how easily the drugs can get into cells; there's an old method for this, but of course when we tried to repeat the Robillard paper, it just didn't work at all. So we had a coffee and complained about our supervisor for half an hour – we also did a rehash of the standard ‘I feel so insecure about the academic path – there are no guarantees – hey, we should sell out and earn four times more for working less hours’ type discussion. Finally we got back to lab and dissolved up the compound; we didn’t have any 1-octanol but the lab next door lent us some, and then we beat our heads against the archaic atomic absorption spectrophotometer the department insists on not replacing. (It actually has an old green monochrome screen!) We loaded the samples and of course the carousel on the AAS doesn’t work, it’s a really crap machine. While it analysed our samples we talked about how we were worried about whether we would ever be able to get our own research funding, and how the people who do get funded don’t seem to do very impressive science. After it printed out our results on a dot matrix printer, we had to re-do half the samples because we didn’t like the numbers it gave us. The second time was much better, so we went to the pub for a while, where we talked about how unfair it is that we aren’t paid as much as corporate people, and how unfair it is that society doesn’t value us. I got my summer student to do the calculations because I couldn’t be bothered and really I don’t remember any of that log stuff from high school.

If only we could tell it like it really is!