OP mentioned drug discovery. If he wants to go into that direction or if he would like to go into protein modelling a good grasp of statistical physics is essential, you can't do without.