Question: What, in detail, is involved in your simulations on superbugs?

Eleanor Turpin answered on 27 Jun 2012:

The simulations themselves are not of the superbugs but of an antibiotic that can kill them.

I always simulate one molecule of the antibiotic and sometimes one molecule of the material that makes up the cell wall of the bacteria to see how they interact. If the antibiotic binds to the cell wall molecule then we know that it’s very likely to kill the bacteria in real life. That’s because the cell wall material will be too busy with the antibiotic to build new cell walls when the bacteria want to multiply.

The simulations themselves are called molecular dynamics because they simulate how molecules move. To start a simulation you need the starting positions of all the atoms in the molecule (which usually someone else has found in an experiment) then you calculate the forces acting on the atoms, move them a tiny bit, calculate the new forces and repeat – perhaps forever. This is a very slow process: it takes me three months in real word time to simulate a one thousandth of a second.

To calculate the forces acting on the atoms is very difficult and sometimes I need to use Quantum Mechanics, which involves some really hard maths.

I sometimes make animations and pictures of my simulations using the same type of software that was used to produce the picture of DNA in your profile picture.

The computer I use is called Jupiter and has 1900 more processors than the computers you use at school. But what makes it really good is that the process can talk to each other 10 times faster than a normal computer. Jupiter is capable of performing over 12 million million calculations per second. It cost over £1 million pounds but has been used to make over £27 million for my university, so was well worth the money.

Thanks for asking about my research. I hope this has helped answer your question but if there is anything more specific you’d like to know about please ask.