Automated Simulation Data Analysis

Pictorial representation of pockets in protein

With the increasingly available computational power the size of simulated data are growing at a pace which is making difficult, or even impossible, the conventional visual inspection process. This calls for novel and efficient techniques aiming at performing a preliminary, almost automated, analysis capable to point the attention of the researcher to the potentially relevant features of a simulation.

In this respect, we developed two approaches that should significantly help the computational scientists analysing their data:

1) Pocketron: a novel tool to analyze overall pocket dynamics along a MD trajectory, pointing at potential allosteric connections between different sites;

2) a novel automated Path Finding algorithm, which finds a regularised path in data space starting from given ends,  in an almost user independent way.