PhD Research Project: Extracting novel algorithms from Nature

The development of algorithms underpins the information processing age and data science. Creating novel algorithms capable for processing information under different conditions and scenarios provides opportunities to extend computation to new domains and under different conditions. A source of inspiration for the development of information processing strategies is nature, and indeed machine learning and genetic algorithms have found support from this domain.

We are used to thinking of evolution acting on physical traits, like strength and speed. However, evolution also acts on algorithms in nature - the rules governing the behaviour that leads to survival and reproductive success. Biology lacks electronic microprocessors and so must implement these algorithms within and using the noisy biochemical environment of the cell. This physical constraint leads to diverse and unfamiliar methods of information processing across life - from noisy gene regulatory networks to signalling cascades and intercellular communication.

The decisions that plants make in response to their environment are arguably the most important for humanity. If plants decide to germinate or flower too soon, too late, or in adverse environments, humans starve. The relative fixed cellular plan of plants means that these decisions have to be made in an even more constrained physical space - plants cannot grow brains. In this sense, computation in plants has evolved to be exceptionally accurate.

We have shown that plants use complex and beautiful biochemical signalling within and between cells to sense, process and integrate temperature variability, to “roll dice” to generate randomness, to enhance photosynthetic capacity, and to “hedge bets” to rationally deal with unpredictable environments. These discoveries have revealed new ways in which life has solved the problem of implementing algorithms to process and respond to large-scale data in a noisy world.

This project will combine biological, mathematical, and computational perspectives to further discover how life implements vital algorithms to process large-scale data in noisy environments. Generalising across organisms, we will identify modes of information processing of cold temperatures in plants, and use modelling and experiments to explore their behaviour and ways they may be rationally engineered to improve organism performance in different environments.

The application of these algorithms to other computational problems will also be explored to determine whether these are of use for emerging technological problems. As technology and data pervade more and more situations in life and the environment, the need for novel information processing solutions is increasing. Some examples are distributed computation across noisy devices in the internet of things; communication and decision making in populations of drones; and decisions emerging from large-scale environmental sensing assays. We anticipate that the evolved algorithms that nature uses to process information in noisy, distributed settings will inform novel solutions to these emerging technological problems.

The ideal student will have some experience with mathematical/physical models and the use of computers in scientific contexts (simulation / modelling / statistics). Importantly, they will be excited about the complexity and ingenuity of biological life and enthusiastic about discovering how natural systems have evolved to deal with the complex challenges of a noisy world.

Funding Notes

To support students the Turing offers a generous tax-free stipend of £20,500 per annum, a travel allowance and conference fund, and tuition fees for a period of 3.5 years.

References

• Topham, A.T., Taylor, R.E., Yan, D., Nambara, E., Johnston, I.G. and Bassel, G.W., 2017. Temperature variability is integrated by a spatially embedded decision-making center to break dormancy in Arabidopsis seeds. Proceedings of the National Academy of Sciences, p.201704745.
• Johnston, I.G. and Bassel, G.W., 2018. Identification of a bet-hedging network motif generating noise in hormone concentrations and germination propensity in Arabidopsis. Journal of The Royal Society Interface, 15(141), p.20180042.
• Williams, B.P., Johnston, I.G., Covshoff, S. and Hibberd, J.M., 2013. Phenotypic landscape inference reveals multiple evolutionary paths to C4 photosynthesis. Elife, 2, p.e00961.