Skip to main content

MSc Mini-projects

Mini-project 1

Pulse-coupled oscillator networks: Do small-worlds synchronise quicker?

My first 12 week mini-project was supervised by Stefan Grosskinsky(University of Warwick) and Marc Timme (Max Planck Institute for Dynamics and Self-Organization). We study the effects of topology on the synchronization time in pulse-coupled networks, looking in particular at small-world networks.


Poster introducing the project and showing some preliminary results.

Mini-project report giving full details and results.

The results from this project contributed to the following publication:

C. Grabow, S. Hill, S. Grosskinsky, and M. Timme (2010)
Europhys. Lett. 90:48002.
 

 

Mini-project 2


An informative pathway-based prior for combinatorial inference with an application to protein-signalling networks

My second 12 week mini-project was supervised by Sach Mukherjee (University of Warwick). In this study we develop an informative prior distribution, allowing existing biological knowledge to be incorporated into Bayesian variable selection. We use the sparse noisy Boolean function model and Bayesian variable selection method developed in Mukherjee et al. (2009). Boolean functions can model combinatorial influences between system components and the response of interest. Our informative prior utilises network and pathway structure, scoring variable subsets based on the notion that variables "close" to each other in a pathway are more likely to be involved in determining a response. We test our approach on synthetic data from two simulation regimes and apply it to proteomic data obtained from a study of signalling in breast cancer, the output of interest being drug response.

S. Mukherjee, S. Pelech, R. M. Neve, W-L. Kuo, S. Ziyad, P. T. Spellman, J. W. Gray and T. P. Speed (2009). Sparse combinatorial inference with an application in cancer biology. Bioinformatics 25(2): 265-271.


Mini-project report giving full details and results.

This work contributed to a recently submitted paper:

Integrating biological knowledge into variable selection: an empirical Bayes approach with application to cancer signalling
S. M. Hill, R. M. Neve, N. Bayani, W-L. Kuo, S. Ziyad, P. T. Spellman, J. W. Gray and S. Mukherjee
submitted for publication