Introduction

Comparative Modelling allows us to build a three-dimensional (3D) model for a protein of known amino acid sequence, but unknown structure using another protein of known sequence and structure as a template.

Comparative Modelling used to be known as "Homology Modelling", implying that models were always generated from homologous proteins. Generally this is still the case. However, techniques such as fold recognition and threading allow us to recognise that a sequence may adopt the same fold as a non-homologous protein and, providing a satisfactory alignment can be achieved, a 3D model can be built.

Aims and Objectives

From this lecture and practical you will (1) gain an understanding of the concepts behind comparative modelling, its powers and limitations; (2) have a feel for the accuracy which may be achieved and the factors which limit the accuracy.

You will be able to:

1. Describe the two major approaches to comparative modelling and give a detailed explanation of one of those approaches,
2. Explain how an RMS deviation is calculated,
3. List the major factors which contribute to the accuracy of a model,
4. Evaluate for a given model which regions are likely to be accurately modelled and which regions will be poor.

Online References

http://www.cmbi.kun.nl/gv/projects/overview/over2.html Links to a number of articles by Gert Vriend on protein modelling.

http://www.cmbi.kun.nl/gv/articles/text/gambling.html "Professional Gambling" A very nice article explaining the concepts of protein modelling