i wanted to know to what extent simulation of biomolecules structure can be accurate and predict true behaviour of these molecules.

I think that this is a rapidly evolving research area. I have seen many instances in which the specific folding pattern of proteins can show why and how a resulting active patch is made and functions in some cellular process. I think that it is only a matter of, relatively short, time to when there will be the computer power to design a protein for a specific purpose, and to then to be able to work back to and produce the gene needed to produce the protein. SM

i wanted to know to what extent simulation of biomolecules structure can be accurate and predict true behaviour of these molecules.

 

It really depends on how much computational power you have and what your goals are. Homology-based programs are not all the computationally demanding, but that's because you've already solved related structures. OTOH, if you need a dynamic simulation over a range of conditions for an unknown protein, you're going to need to rent some computational cycles from something like Folding@home or Rosetta@home.