Dr Chris Knight
The genotype-phenotype map in evolution
How do the individual DNA changes used by evolution (genotype) affect the behaviour of the complex system that is the living cell (phenotype)? How do organisms manage to evolve at all when even a small genetic change may affect many different aspects of the cellular system? How do the DNA changes used by short-term evolution relate to the DNA changes seen across longer-term evolution across strains or species?
To explore the answers to such questions I use a variety of approaches including:
* microbial experimental evolution, combining the power of 'model' genetic systems and with “'omic” technologies for measuring many aspects of phenotype at the same time.
* Bioinformatic analysis of fully sequenced genomes to look at sequence evolution across longer time-scales.
* in silico evolution where DNA sequences evolved in a computer are tested experimentally, looking at the genotype-phenotype map in such an abstracted system makes it possible to control all aspects of the evolution and look at many thousands of DNA sequences.
All of the above requires appropriate ways to abstract meaningful insights from complex biological systems. This requires mathematical or statistical models. What is the appropriate level of complexity for such models? How can the results of very complex models be analysed, visualised and interpreted meaningfully? These questions cut across all my work, but particularly concern:
* Growth curves: We can monitor the growth of a microbial population in great detail, but then we need to use that data effectively. In collaboration with a worm laboratory, we are also applying our approaches to microbial 'death' curves as they are eaten by nematode worms.
* Mutation rate control: we are involved in testing models developed by mathematicians about how mutation rate may vary in evolution.
* Systems biology models: We are using mathematical models of the biochemical details of the cell (both signalling and metabolism) to ask how these systems evolve.
* climate models: We have been involved in analysing data from the climateprediction.net project to see how uncertain or variable aspects of the model affect outputs relevant to predicting climate change.