Modelling the Evolution of Molecular Components, Systems & Behaviours in respect of the transition of Dr Richard Goldstein from MRC NIMR to UCL

Evolution is important for three reasons. Firstly, understanding how evolution occurs is inherently critical to many processes that involve living objects. For example, some pathogens (e.g. RNA viruses) evolve in timescales of weeks to years; combating these public health threats require understanding how these threats change and shift due to the evolutonary process. Secondly, the evolutionary record provies a wealth of information about structure, function, and physiological roles. Decyphering this record requires an understanding of the process that resulted in the observed variety. Thirdly, evolution is the essence behind all biological phenomenon. If we wish to understand why things are the way they are, how the process of evolution determines the resulting properties of biological components, systems, and organisms, and how these properties affect the biological system, we need to understand this evolutionary process. We are modelling a number of different evolutionary processes, including evolution of viruses (influenza, papillomavirus), protein structures, and biological networks. We are in particular interested in how these processes can change, for instance, when a virus encounters a new host, or when the needs of the network are different. It is hoped that these models will allow us to understand the evolutionary process better, providing important understanding of existent biological systems. The work with viruses, in particular, is important in understanding and monitoring emerging health threats.

We are working in a variety of areas in understanding the evolutionary process, modelling this process, and applying these models to specific biological and medical problems. 1) Evolution of pathogens We are modelling the molecular evolution of influenza, both at the DNA and protein levels, a collaboration with Alan Hay (virology) and other members of the EU FLUPOL project. Most evolutionary models assume that the evolutionary process is similar for all locations in these biomolecules at all time. We are interested in moving beyond these models, including temporal and spatial heterogeneity. In particular, we are interested in looking at how the process changes when influenza switches hosts. This has allowed us, for instance, to reconstruct the history of the 1918 Spanish flu epidemic, as well as identifying protein locations throughout the influenza genome where the selective constraints are different in avian and human viruses, work that will be extended to include swine viruses. These analyses can provide important information to better direct the widespread virus surveillance projects, as well as allowing us to better interpret the results of these efforts. We have also looked at the evolution of papillomavirus, a collaboration with John Doorbar (virology), developing methods to determine how this virus spread over the wide range of hosts. These projects have involved the construction of new, generally useful techniques to understand and model molecular evolution. 2) Models for better phylogenetics Most models of sequence change only consider changes in DNA or amino acid, neglecting the process of insertions and deletions. This is because there are no good models of this process that have been included in phylogenetic reconstruction methods. In collaboration with Simon Wheelan (Manchester), we are developing reconstruction models that include insertions and deletions, modelled using a Tree-Based Hidden Markov Model (T-HMM) formalism. 3) Evolution of chemotaxis In collaboration with Orkun Soyer (CoSBI, Trento) we are working to understand how biochemical networks can evolve to mediate simple behaviour, in our case chemotaxis. We are looking at the relationship between the evolutionary context (including the environment, the available biochemistry, and the mechanics of evolutionary change) and the chemotaxis process that results. This work is directed towards understanding both how evolution is able to evolve networks, as well as understanding the chemotaxis process itself, especially the variety of possible strategies that might exist beyond the well-studied E. coli model system. 4) Evolution of protein structure Most models of protein evolution consider the amino acid or DNA sequence. Much less is known about the process of structural change. In collaboration with Willie Taylor (Mathematical Biology) and Jotun Hein (Oxford) we are developing models of this structural change. It is hoped that this work can look deeper into the past, as this structural change occurs at a much slower rate than sequence change. It can also provide insight into the process that gave us the observed universe of protein structures.