For MR solutions with more than 4 molecules in the ASU, you can improve your chances of finding a solution by employing the following considerations:
The number of protein units in the ASU predicted by Matthew's analysis my not be accurate for n>4. Always look at the molecular packing of potential solutions in the ASU using Coot or Pymol. You may discover you have too many molecules, and therefore overlaps, or too few molecules, leaving protein size gaps in the crystal lattice. Lattice packing of partial solutions may guide your thinking about the correct number of molecules in the ASU. A good solution should normally have symmetrical and clear solvent channels, and good contacts between all molecules.
You may nave a better chance of obtaining a initial solution by searching with dimers (if structurally appropriate) than monomers.
Unless your crystal form is identical to the solved version, there is no guarantee your N will be the same as the solved N molecules in the ASU. Use the Matthew's analysis and lattice packing as your guide.
Using this approach I've been able to obtain MR solutions for ASUs with N=6-8, even with identities as low as 30%.
Good luck.
Roger Rowlett
Gordon & Dorothy Kline Professor, Emeritus