Chris Lovell

Event-B code generation: type extension with theories

The Event-B method is a formal modelling approach; our interest is the final step, of generating code for concurrent programs, from Event-B. Our Tasking Event-B tool integrates Event-B to facilitate code generation. The theory plug-in allows mathematical extensions to be added to an Event-B development. When working at the implementation level we need to consider how to translate the newly added types and operators into code. In this paper, we augment the theory plug-in, by adding a Translation Rules section to the tool. This enables us to define translation rules that map Event-B formulas to code. We illustrate the approach using a small case study, where we add a theory of arrays, and specify translation rules for generating Ada code.

An artificial experimenter for enzymatic response characterisation

Identifying the characteristics of biological systems through physical experimentation, is restricted by the resources available, which are limited in comparison to the size of the parameter spaces being investigated. New tools are required to assist scientists in the effective characterisation of such behaviours. By combining artificial intelligence techniques for active experiment selection, with a microfluidic experimentation platform that reduces the volumes of reactants required per experiment, a fully autonomous experimentation machine is in development to assist biological response characterisation. Part of this machine, an artificial experimenter, has been designed that automatically proposes hypotheses, then determines experiments to test those hypotheses and explore the parameter space. Using a multiple hypotheses approach that allows for representative models of response behaviours to be produced with few observations, the artificial experimenter has been employed in a laboratory setting, where it selected experiments for a human scientist to perform, to investigate the optical absorbance properties of NADH.

Organising Chemical Reaction Networks in Space and Time with Microfluidics

Information processing is essential for any lifeform to maintain its organisation despite continuous entropic disturbance. Macromolecules provide the ubiquitous underlying substrate on which nature implements information processing and have also come into focus for technical applications. There are two distinct approaches to the use of molecules for computing. Molecules can be employed to mimic the logic switches of conventional computers or they can be used in a way that exploits the complex functionality offered by a molecular computing substrate. Prerequisite to the latter is a mapping of the versatile means to achieve this. In the present paper we review microfluidic technology as a versatile means to achieve this, show how we use it, and provide proven recipes for its application.

Characterising enzymes for information processing: towards an artificial experimenter

The information processing capabilities of many proteins are currently unexplored. The complexities and high dimensional parameter spaces make their investigation impractical. Difficulties arise as limited resources prevent intensive experimentation to identify repeatable behaviours. To assist in this exploration, computational techniques can be applied to efficiently search the space and automatically generate probable response behaviours. Here an artificial experimenter is discussed that aims to mimic the abilities of a successful human experimenter, using multiple hypotheses to cope with the small number of observations practicable. Coupling this approach with a lab-on-chip platform currently in development, we seek to create an autonomous experimentation machine capable of enzyme characterisation, which can be used as a tool for developing enzymatic computing.

Characterising enzymes for information processing: microfluidics for autonomous experimentation

Information processing within biological systems relies upon the interactions of numerous protein macromolecules with one another, and with their environment [1]. Recognising this, enzymes have been applied in the implementation of Boolean logic gates. However, given the structural complexity of enzymes, it would appear that enzyme behaviour is not limited to simple Boolean logic behaviour.

Towards improved theoretical problems for autonomous discovery

Active learning and experimental data acquisition address the same problems, understanding a system under investigation with as few resources as possible. However there are few instances where the theoretically principled techniques in active learning or sequential experimental design have been applied to managing data acquisition in physical experimentation. Partly this is due to fundamental differences between the problems investigated within active learning and the issues faced in much physical experimentation. From a previous study we conducted into autonomous experimentation, where we developed a system capable of automatically designing experiments and proposing potential hypotheses, we aim to investigate and highlight the differences between theoretical active learning and the requirements of experimentalists. We also propose an update of the multi-armed bandit problem that provides a theoretical problem more closely aligned to that found in physical experimentation. We believe that for active learning techniques to be used more widely as tools within physical experimentation, a greater focus of research has to be placed on theoretical problems that have assumptions more closely aligned to those found commonly within physical experimentation. Assumptions such as extremely limited resources, more so than typically considered in active learning problems, along with erroneous observations or noisy oracles, should become standard features of active learning problems, as in experimentation there are rarely enough resources available to be certain about the validity of the data obtained and the quality of the hypotheses produced.

Enabling the discovery of computational characteristics of enzyme dynamics

Biology demonstrates powerful information processing capabilities. Of particular interest are enzymes, which process information in highly complex dynamic environments. Exploring the information processing characteristics of an enzyme by selectively altering its environment may lead to the discovery of new modes of computation. The physical experiments required to perform such exploration are combinatorial in nature. Thus resource consumption, both time and money, poses major limiting factors on any exploratory work. New tools are required to mitigate these factors. One such tool is lab-on-chip based autonomous experimentation system, where a microfluidic experimentation platform is driven by machine learning algorithms. The lab-on-chip approach provides an automated platform that can perform complex protocols, which is also capable of reducing the resource cost of experimentation. The machine learning algorithms provide intelligent experiment selection that reduces the number of experiments required for discovery. Here we discuss development of the experimentation platform and machine learning software that will lead to fully autonomous characterisation of enzymes.