Analysis, storage, and transfer of molecular dynamic trajectories are becoming the bottleneck of computer simulations. In this paper we discuss different approaches for data mining and data processing of huge trajectory files generated from molecular dynamic simulations of nucleic acids.