Quantitative methodologies need constant development to meet the demands from science and opportunities from new computing technologies. BioSS's research is structured in three themes.
Developments in molecular genetics technologies are generating enormous quantities of data, often of new data types, allowing deeper studies of genomes and the relationship between genetics and biological function. Simultaneously, new computing technologies are allowing easy access to rapidly increasing computer processing power and data storage capacity. BioSS aims to develop and automate methodology for analysing these data, harnessing the computing power to extract maximum information from the data.
Mathematical modelling plays a key role in achieving many scientific objectives. BioSS aims to enhance this role by addressing generic issues including: simplification, analysis and approximation of models for complex systems; parameter estimation and model selection in stochastic process models; Bayesian methods for decision support; and methodologies for estimating risks in complex interacting systems. The strategy will be to develop methodology in the context of specific collaborative applications.
Statistical methodology needs constant development, firstly to keep pace with the requirements of new technologies being used in the biological and environmental sciences, and secondly to address new questions that arise as science becomes ever more quantitative. In particular, there is a pressing need for new methodology to correctly interpret large, highly-structured data sets. BioSS will develop and adapt methodology in the key areas of image analysis and spatially-, temporally- and spatiotemporally- structured data.