Professor Kevin Cowtan

01904 328271

Automated protein structure solution

X-ray crystallography is an important tool in the biochemical and biological sciences for the determination of the 3-dimensional atomic structure of biological molecules. This in turn contributes to the understanding of the function of the molecule, its regulation, and its behaviour within a complex biological system.

The structure solution process involves a number of stages, starting with the laboratory based step of crystallisation and data collection, followed by a number of computational steps involved in the computation of the final atomic model from the X-ray diffraction pattern, via an intermediate stage of a 3-dimensional electron density map. My particular interests are the application of image processing techniques to reduce the noise level in the electron density maps, followed by automated interpretation of protein electron density map in terms of atomic coordinates. I also work with Dr Paul Emsley in the development of new software to visualise, manipulate and validate atomic models. 

Video demonstration of automated building of nucleotide chains in an electron density map. (This video contains no sound.)

I have authored several high profile crystallographic software packages, including the "dm" phase improvement software, a suite of crystallographic software libraries ("clipper"), a statistical phase improvement package ("pirate"), and a statistical model building package ("buccaneer"). All of these are distributed through the "CCP4" suite of crystallographic software. I also contribute to the "Coot" molecular graphics software of Dr Paul Emsley. I am also involved in training new crystallographers, both through web-based teaching tools, and at international workshops. My recent work has been supported by a Royal Society fellowship (2000-2008), BBSRC grant (2008-2013) and a CCP4 fellowship (2013 on).

I am also developing methods for the automated interpretation of Electron Microscopy (EM) maps. These present different challenges to X-ray crystallographic maps, typically being lower resolution but less noisy. New refinement methods with do not rely on being able to see atomic features will enable us to interpret these maps.


Climate Science

My climate science research focuses primarily on problems which are relevant to the public understanding of climate science. With my colleague Robert Way I have been investigating biases in historical temperature record from weather stations. Our primary work concerns temperature change over the past two decades. The main temperature record providers show a slowdown in the rate of warming over this period, however when biases in the temperature record are taken into account, we find that part of the slowdown disappears.

Video explanation of the impact of coverage bias on the recent temperature record.

A further contribution to the apparent slowdown arises from a change in the way sea surface temperatures are measured, explained by my colleage Zeke Hausfather of UC Berkeley.

Video explanation of the impact of sea surface temperature observation methods on the recent temperature record.

I am also involved in climate science communication and contribute to a massive online course run by the University of Queensland. I can offer undergraduate and postgraduate projects for students who are interested in climate change, data analysis, science communication or science denial.