Exploring the function and evolution of proteins using domain families

• Main Author: Reid, A. J.
• Published: University College London (University of London) 2009
• Subjects: 572.6
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.564630

Proteins are frequently composed of multiple domains which fold independently. These are often evolutionarily distinct units which can be adapted and reused in other proteins. The classification of protein domains into evolutionary families facilitates the study of their evolution and function. In this thesis such classifications are used firstly to examine methods for identifying evolutionary relationships (homology) between protein domains. Secondly a specific approach for predicting their function is developed. Lastly they are used in studying the evolution of protein complexes. Tools for identifying evolutionary relationships between proteins are central to computational biology. They aid in classifying families of proteins, giving clues about the function of proteins and the study of molecular evolution. The first chapter of this thesis concerns the effectiveness of cutting edge methods in identifying evolutionary relationships between protein domains. The identification of evolutionary relationships between proteins can give clues as to their function. The second chapter of this thesis concerns the development of a method to identify proteins involved in the same biological process. This method is based on the concept of domain fusion whereby pairs of proteins from one organism with a concerted function are sometimes found fused into single proteins in a different organism. Using protein domain classifications it is possible to identify these relationships. Most proteins do not act in isolation but carry out their function by binding to other proteins in complexes; little is understood about the evolution of such complexes. In the third chapter of this thesis the evolution of complexes is examined in two representative model organisms using protein domain families. In this work, protein domain superfamilies allow distantly related parts of complexes to be identified in order to determine how homologous units are reused.