During these last years, modelling in physics remained blind to the advances of sisters disciplines, especially mechanics and robotics, that have developed powerful tools in rigid body dynamics, soft body dynamics, and fluid dynamics. Techniques as inverse kinematic and physics engines (or game engines) are nowadays commonly used in video games for very realistic results. I show here how it is possible to simulate assemblies of DNA and proteins (chromatin fibres, and, at larger scale, chromosomes) as articulated system, by means of physics engines. I also show how to replace the standard local Langevin thermostat by a global one that achieves a much faster sampling of the configuration space of the system (canonical ensemble). This new thermostat results to be of particular interest when used with physics engines. The simulations that I have performed with these tools well reproduce magnetic tweezers experimental results and allows to make predictions. Finally, this simulation technique offers interesting perspectives for the modelling of whole cell nuclei for different organisms, ranging from Drosophila to humans, and provides therefore a precious tool for the interpretation of the most recent experimental data on genome architecture.
