Our research area is Computational Biology with particular emphasis on the application of biomedical engineering and applied maths to physiology and molecular biology. Currently our research focuses on three areas:

• Synthetic Biology

• The aim of this proposal is to engineer a synthetic biological network for in vivo regular therapeutic delivery of insulin in a rhythm corresponding to normal nutrient uptake. To this end, we will engineer stable synthetic ÒoscillatorÓ networks in yeast and mammalian systems able to express mRNA/protein levels with a predetermined frequency and amplitude. The synthetic oscillator network has to guarantee stable and synchronised oscillation in the cell population. The yeast will be used as a Òtest-bedÓ for the synthetic biology design strategies developed in this project. In the context of the mammalian tissue, individual cellular oscillators have to be synchronised in order to fulfil the macroscopic function of an insulin delivery device. Hence, the engineering of the synthetic network involves additional inputs and outputs that enable resetting of the oscillators. In view of therapeutic applications, the desirable system would reset insulin oscillations with the circadian rhythm. Specifically, the synthetic oscillator in the mammalian system will be connected to circadian signals like PER 1 and CRY. To achieve this aim, COBIOS brings together scientists from yeast and mammalian molecular biology, computer science, engineering and control theory. We will employ methods from systems dynamics and control theory to develop and implement modular control networks that enable oscillations in the networks they will be connected to. In particular, we will address the problems of (i) robustness of controller dynamics, (ii) suitable interfaces to the controlled networks, and (iii) mechanisms for regulation of the controllerÕs dynamics characteristics (e.g. period and amplitude) through external signals that can be exogenous (yeast system) or outputs of cellular signal processing (the circadian clock in mammalian systems) at the levels of individual cells and tissues.

• Reverse Engineering and modelling of complex gene regulatory networks

  • The identification of gene regulatory networks is of major importance in order to understand the working mechanisms of the cell in patho-physiological conditions. Mathematical models of gene-networks would also help in the optimization of novel drugs.

  • Our research has focused on the development of experimental protocols and computational algorithms to detect gene regulatory network using a small number of experiments. The algorithm we developed [Gardner TS, di Bernardo D, et al. Science, 2003] uses measurements of mRNA concentration of the genes that are part of the network. These measurements are repeated in a number of experiments that varies with the size of the network. In each experiment one of the genes of the network is perturbed. The perturbation means simply the ÔupregulationÕ of the gene. This is achieved in prokaryotes with plasmids having a copy of the coding sequence of the gene that must be perturbed, and controlled by a promoter whose efficiency can be modified by the experimenter. From these measurements it is possible to reconstruct the causal relationships among the different genes, and therefore the regulatory network. We also showed that it is possible to use reverse engineering methods to infer a drug mode of action from gene expression alone on a genome wide scale. We developed a method named MNI (Microarray Network Identification) [di Bernardo D, et al., Nature Biotechnology, 2005] . MNI is able in yeast to build a model of a gene regulatory network in from microarray data and use the model to filter gene expression data in response to drug treatment and select the genes directly interacting with the drug. We are now extending our approach to infer the local regulatory network in which a gene of interest is embedded. Our aim is to infer a model of the regulatory network from the analysis of gene expression profiles measured a consecutive time intervals following the perturbation of the gene of interest.

• DNA sequence analysis

  • It has been recently discovered that RNA not coding for any protein ( ncRNAs ) have an important functional role in the regulation of transcription and translation in the cell. No computational tools are yet available to automatically detect ncRNAs in the genomic sequence. We have developed an algorithm [di Bernardo D, et al. Bioinformatics, 2003] that uses alignments of synthenic sequences in different genomes to detect ncRNAs . This algorithm can detect ncRNAs in multiple alignments. Its computational complexity is proportional to the square of the sequence length, and thus makes it suitable for whole genome scans.

  • We are now focussing on the analysis of transcription factor binding sites (TFBS) in multiple organisms. We are developing an algorithm that makes use of a probabilistic approach to compare promoters from orthologous genes and select functionally active binding sites. We want to understand how the TFBS evolved and to what extent they have being conserved in higher organisms.