Archive for October, 2010

Comment on “Publish your computer code: it is good enough”

Monday, October 18th, 2010

One essay published in Science. The author, Nick Barnes definitely has a point, scientists need publish their computer codes, it is not only in the interests of others but also their own. Needless to say, it is an explosive idea to many people, one indication is that there are many comments online after the publication. So I add my trivial here.

This is indeed a good call. But I also have some minor concerns of the effectiveness of publishing codes. Scientists are supposed to work on solving problems (or finding the right problems and/or solving them). That is the motivation for them, so they are highly reluctant to spend time and effort on something that can’t produce immediate good in the perspective of science for themselves. Also given the harsh environment, they are under high pressure of producing papers,¬† not codes, it is hard to convince them unless they are forced to do so. So published codes might not get good attention at all. Another concern is the quality of published codes, in particular, given the fact that lots of scientists are not well trained to be software engineers. Doing science means lots of try and error (well, genius are excepted). The final code might evolve several generations, so it is a collection of several shaking but working code. After the paper is accepted, it would cause a lot to recall what the code means not because it is not understandable anymore but his/her mind is fully occupied with other potentially exciting project(s). So if just publish the last version, it would be hard for anybody to quickly understand the code. Look at the code like this for a while, I’d wonder if worth just write my own, which is “plain and very easy to understand to anyone”(I, as an author surely hope so :( ).

Nevertheless, maybe some lone men at the corner of an office want the codes regardless, if for that sake, just publish it.

Good obituary

Sunday, October 3rd, 2010

The obituary is about George C. Williams by Richard Dawkin and published in Science. The first two sentences are very good and I can’t help posting here:

It has become a clich√© that Charles Darwin would not have succeeded as a scientist today. He would not have won big research grants and did not have the mathematics to be a “theorist” by today’s conventions.

Well said, well said. From now one, I shouldn’t feel bad myself at all, even that poor Charles Darwin can’t get a grant, why should I expect?

Protein-ligand binding equlibria

Friday, October 1st, 2010

Reversible binding of ligands to proteins is similar to reversible folding of proteins and involves reaching equilibrium. The binding reaction is

Protein \cdot Ligand \leftrightharpoons Protein + Ligand

The dissociation constant which describes the equlibria is

 K_{D} = \frac{[Protein][Ligand]}{[Protein \cdot Ligand]}=\frac{[P][L]}{[P\cdot L]c^{0}}, \ with\ c^{0}=1 mol\ L^{-1}

So when it is reported K_{D}=10^{-9}mol\ L^{-1}, it should be just K_{D}=10^{-9}.

And Gibbs free energy change of the binding is related to the dissociation constant:

\Delta G^{binding} = \Delta H - T \Delta S = -RT\ lnK_{D}

A good reference of this post is from “Introduction to Protein Science: Artchitecture, Function, and Genomics by Arthur M. Lesk”.