I used to work in aeronautics at NASA Ames Research Center. I worked in a big wind tunnel building, and the people I worked with were either wind tunnel engineers or CFD engineers. CFD stands for Computational Fluid Dynamics. They’re the people who write programs to simulate the flow of air around an airplane. They always had the prettiest pictures to show management. It didn’t matter if the simulations were inaccurate, they were beautiful and convincing. This drove the wind tunnel guys crazy. Their wind tunnels were ugly, loud, and incredibly expensive to run, but often they were the only way to get an answer you could trust. They were the Rodney Dangerfields of the center.
But simulation has come a long way in the last decade. People sometimes speak of three pillars supporting scientific progress as being experiment, theory, and simulation. Simulation, once the weak sibling of the three, has seen fantastic growth in power and scope. In molecular biology, like aeronautics, experiments are difficult, slow, and expensive. You do as many as you can afford, but sometimes simulation is the only way to get the information you need.
Seed magazine has produced a nice video describing the nature of science in silico, which is to say, science as simulated on a silicon computer chip. Follow the link to get to the video: Science In Silico
Rambles Blog at Star Chamber 
