Molecular systems biology

Molecular systems biology

We are interested in the general paradigm of systems biology as it is applied to mechanistic models of protein-protein interactions. We have several local collaborations with experimental laboratories at the Sir William Dunn School of Pathology.

Please contact Dr. Omer Dushek, Dr. Ruth Baker, Dr. Eaamon Gaffney, or Prof. Philip Maini for more details.

Systems biology of lymphocyte signaling

The major challenge in molecular immunology is no longer to identify new molecules but to understand the interaction networks that the identified molecules form. We are directly addressing this challenge by using quantitative molecular and cellular experiments combined with mathematical modeling to understand the complex signaling networks underlying lymphocyte activation with the aim of exploiting this knowledge for the benefit of human health.

In collaboration with Dr. Omer Dushek (

Systems biology of fibril formation with applications to Alzheimer's disease.

The pathological consequences of protein misfolding, aggregation, and amyloid deposition are a hallmark of a wide variety of clinically unrelated diseases, such as type II diabetes mellitus, Alzheimer’s disease, and prion diseases. These characteristic tissue deposits contain fibrils originating from a common process of nucleation-dependent polymerization consisting of an energetically unfavorable nucleation step followed by the formation of a minimal self-assembled complex (stable nucleus) that further elongates to produce a final fibrilar assembly. In turn, this multistep amyloid aggregation can lead to a common path of cell membrane disruption, interference with cellular processes, and apoptosis. The process of fibrilization remains poorly understood. We aim to understand the mechanism of fibrilization and to develop novel drugs to perturb the process to achieve disease free states.

The mechanism of fibril formation is routinely investigated by examining fibrilization kinetics. In these experiments, an initial concentration of pure monomers begins to form oligomers that eventually form stable nuclei from which fibrils can elongate. However, it is only possible to observe fibril elongation and no direct information is available on the intermediate states and hence the path of fibrilization. Reverse engineering the fibrilization pathway is possible by using systems biology models that include multiple proposed pathways. Comparisons between model predictions and experiments can then be used to determine plausible pathways. Ultimately, mathematical models can be used to investigate the effects of drugs and can propose novel targets that can disrupt fibrilization.



Figure Caption: Kinetics of fibrilization for islet amyloid polypeptide (IAPP). Fluorescence intensity (y-axis) measures the amount of IAPP incorporated into fibrils as a function of time (x-axis) for several starting concentrations of monomeric IAPP.

In collaboration with David Vaux (

Systems biology of centrosomal proteins

Centrosomes are an important subcellular organelle with functions in cell division and cell polarity. They consist of a pair of protein cylinders called centrioles surrounded by a dense matrix of various proteins collectively referred to as the Percentriole Material or 'PCM'. The organisation of the PCM is key to the proper function of centrosomes. The focus of our work is to achieve a systems-level understanding of the many protein interactions that make up the PCM.

In the laboratory of Dr. Jordan Raff at the Sir William Dunn School of Pathology (University of Oxford) fluorescence microscopy techniques, such as fluorescence recovery after photobleaching (FRAP), are routinely used study PCM proteins. These microscopy techniques provide quantitative information on the diffusion, transport, and binding of various PCM proteins. Mechanistic mathematical models are then fit to the wealth of quantitative microscopy data to infer specific protein dynamics.

Most recently, we have been studying the dynamics of a protein known as Centrosomin (Cnn). This protein shows peculiar FRAP dynamics. Unlike most protein, Cnn recovery begins in the centre of the centrosome and then spread outwards (see figure). Using a set of PDEs and ODEs, we have been investigating plausible models that can explain the peculiar behaviour of Cnn. Hypotheses generated by these models are under experimental investigation. Repeating this approach for other PCM proteins we hope to construct a systems model that incorporates the dynamics of multiple PCM proteins. The model can then be used to study, for example, the centrosome maturation pathway.


Figure Caption: A) Schematic of centrosomes. B) Experimental observation of FRAP evolution of Cnn fluorescence (upper panels). Time T0 denotes pre-bleach state and times T1 to T11 denote post-bleach recovery states. Bottom panels depict the corresponding states predicted in silico by our model.

In collaboration with Dr. Jordan Raff (