Latest work Multisomes Protein complexes DNA oscillator Network analysis

Latest work

My latest work cannot be discussed here yet because it is under consideration for publication. Please contact me if you would like to learn about it.

Project supervised by Hagan Bayley (Department of Chemistry, University of Oxford).

Villar, G., Graham, A.D. and Bayley, H. (2012). Submitted.


Aqueous droplets made in oil can be joined by lipid bilayers, which mimic the membranes of biological cells. Networks of droplets connected in this way can use membrane proteins to function as light sensors, batteries and electrical circuits. However, droplet networks were previously confined to a bulk oil phase, which precluded interaction of the droplets with physiological environments.

We showed that droplet networks can be stabilized in water by encapsulating them in small drops of oil to form multisomes. The droplets could communicate with each other and their environment through membrane pores, and could release their contents upon a change in pH or temperature, with potential medical applications. This project involved the development of an encapsulation process, a novel probe for electrical recording, physical modelling, and the application of a mechanism for pH and temperature sensitivity.

Project supervised by Hagan Bayley (Department of Chemistry, University of Oxford).

Villar, G., Heron, A.J. and Bayley, H. Formation of droplet networks that function in aqueous environments. Nature Nanotechnol. 6, 803–808 (2011).
Needham, D. Lipid structures: a brief history of multisomes. Nature Nanotechnol. 6, 761–762 (2011).
Eisenstein, M. Building better bubbles. Nat. Methods. 9, 13 (2012).
Villar, G. and Bayley, H. “Functional droplet interface bilayers.” In Encyclopedia of Biophysics (Springer 2012).

Self-assembly and evolution of protein complexes

I used a simple model of a protein to study the self-assembly and evolution of homomeric protein complexes. This involved an analytical derivation of the thermodynamics of the system, and a study of the kinetics through Monte Carlo simulations.

Project supervised by Jonathan Doye and Ard Louis (Depts of Chemistry and Physics, University of Oxford).

Villar, G., Wilber, A.W., Williamson, A.J., Thiara, P., Doye, J.P.K., Louis, A.A., Jochum, M.N., Lewis, A.C.F. and Levy, E.D. Self-assembly and evolution of homomeric protein complexes. Phys. Rev. Lett. 102, 118106 (2009).

Design of a DNA oscillator

This project aimed to find DNA sequences that interact to form an oscillating system. I formulated the problem as a set of coupled ordinary differential equations, and used linear stability analysis to determine whether given conditions yielded oscillations. This analysis gave guidelines for the design of the DNA strands. I then wrote a computational evolutionary algorithm to optimize a DNA sequence for the given design pattern, and performed experiments to test the predictions of the algorithm.

Project supervised by Andrew Turberfield (Department of Physics, University of Oxford).

Villar, G. Evolving a DNA oscillator. Poster presented at the Doctoral Training Centre Annual Conference, University of Oxford (2010).

Systematic network analysis

I compiled a large database of real and synthetic networks, together with summary statistics of each network. I used this database to perform systematic model fitting, empirical feature selection and classification of both networks and metrics, and to analyze network evolution and damage. This project involved computational implementation of network analysis and machine learning algorithms, and the effective visual representation of complex data.

Project supervised by Nick Jones (Department of Physics, University of Oxford).

Villar, G., Agarwal, S. and Jones, N.S. High throughput network analysis. Proceedings of the Workshop on Analysis of Complex Networks, ECML-PKDD (2010).