Since you didn't get a reply so far, a few random remarks of mine:
- You're very vague with "an MD simulation with atoms in water". There is a huge amount of possibilities for you. The most obvious one, although not directly answering your question, is simulating a smaller system. That may sound disappointing, but from my experience you're often better off with a smaller system that you can simulate properly rather than a larger system whose results you cannot really trust because sanity checks would take too long to perform.
- I'm not a super expert on MD simulations of ?, but the first thing that comes to my mind when thinking about simulating water is its polarization that the LJ fluid doesn't have - I'd at least add an electric dipole to your water beads if you need it.
- Another question you should ask yourself is: why do you want water in your system? If you only want it because your system is in water in reality, but you don't expect the water to have any effect then you should he honest about this point and throw out the water altogether rather than adding something to your simulation and pretending it is water.
- There is a lot of standard MD software available. You may check them out whether they are an option for you, or maybe just how they simulate coarse grained water.