Xiaofan at GNER loco 'Cancer Research UK', Leeds Station

Research interests

My research goal is to model and simulate the behaviours of biological molecules and their interactions as real as possible. Working mainly on atomic resolutions I use Brookhaven PDB entries for molecular modelling, and let them, as objects, play in a crowded environment which truly accommodates the electrostatic and hydrophobic interactions that result in newtonian forces and torques exerted on players.

With the outcome of such simulations we could possibly extract richer information on protein-protein interactions. Traditional pairwise algorithms, such as docking or sequence pattern recognition, have a static nature, which cannot tell when, where and to what extent can a protein interaction take place; however by extending the investigation by timing, populational and environmental factors we will have clearer views of protein interactions closer to their native procedures of occurence.

If the numerical measurements of those mock interactions can be experimentally verified, we may then be able to generate intensive analysis with regard to a perticular protein interaction (e.g. pharmaceutically interested) under all conditions that are not do-able in an experimental lab due to difficulties in technology and costs. This simulated screening process can eventually help people design and evaluate better drug candidates.

My current biological targets include proteins involved in signalling pathways.

Current project

To do what I have proposed above needs a fine-tuned application framework, while the BioSimz library set up by me aims at exactly that. It's an object-oriented, generic-typed C++ library modelling and simulating protein dynamic behaviours at their atomic details. It is under active development currently and you can have a peek at the recent development debriefs and docs here.

Educational qualifications

  • BSc in Biochemistry and Molecular Biology (2005), Shandong University, China
  • MPhil in Computational Biology (2006), Cambridge University, United Kingdom

    • Past project experiences

  • ANN-based pattern recognition of MHC I/II binding peptides, 2005, Shandong University
  • MarkovPack, a C++ package of classical hidden Markov models (HMMs), 2006, Cambridge University
  • humAb: The murine monoclonal antibody humanisation system, 2006, GlaxoSmithKline Medicines Research Centre

    • Publications

  • Not yet (although this nature paper (not by me) can be some healthy salt to read)

    • Hobbies

  • Flight/train/car simulator games (my work benefited quite much from those)
  • Everything digital, computing, networked or programmable
  • Food, as long as it's not English

    •  

    graphical counter