- Join us
Member of Clare Hall
PhD student in Dr Huggins's group
Office: 536 Mott Bld
Phone: +44(0)1223 3 37456
Email: bwji2 @ cam.ac.uk
TCM Group, Cavendish Laboratory
19 JJ Thomson Avenue,
Cambridge, CB3 0HE UK.
Improving Solvation Entropy Convergence:
I am researching techniques to improve the convergence and accuracy of solvation entropy estimates in the inhomogeneous fluid solvation theory (IFST) framework.
Currently a non-parametric 'k-nearest neighbours' (KNN) algorithm is used to achieve an unbiased asymptotic convergence to the entropy difference between a bulk fluid and a fluid with a solute.
For a high throughput analysis of solutes, it is necessary to maximise the rate of convergence of this algorithm in order to minimise the required length of the input trajectory data.
I am currently investigating the implementation of a higher order version of the KNN algorithm which takes three body entropy contributions into account.
In Plain English
I am working on a method to add accuracy to a framework called Inhomogeneous Fluid Solvation Theory (IFST). The method can predict how favourable it would be to add something (an atom, a molecule, a protein...) into a solution. The current way we measure the relavant quantities is guarenteed to get better with more data. The theory currently only looks at interactions betten the solute and the solvent. I'm working on a way to include additional information into the estimate to take into account the solvent interacting with itself given that a solute is present.