I am a small molecule crystallographer currently working with a protein crystal using CCP4, so a bit outside my comfort zone.

I am looking at a protein + ligand structure. The protein without ligand has been published before. There is also a similar protein + same ligand in the pdb, with a slightly different sequence. I expect the packing of my protein + ligand to be similar to this last one mentioned, despite the difference in sequence, P2(1) with Z' = 2 (whereas the protein on its own crystalizes in P2(1)2(1)2(1) with Z' = 1 ).

If I use sequence and .pdb files from the protein without ligand, I end up with Z' = 1 and a lot of empty space in the unit cell, which makes it very difficult to place the ligand with Coot and refine it.

How can I tell Phaser that the solution file must contain 2 chains (Z' = 2)? I tried loading the .fasta file twice in "define unit cell content", but then if I run this sequence with the protein .pdb file on Phaser I get an error message and the job doesn't run.

If I use sequence and .pdb files from the slightly different protein with same ligand, I end up with Z' = 2 but very bad refinement parameters, probably because the sequence is different. Also the solution already contains the ligand in it.

Wouldn't it be better to place and refine it myself instead?

Is there a way I can combine the sequence from the protein file on it's own with the .pdb file of the slightly different protein + ligand in CCP4?