I've been working on the problem I posted earlier, in which Modeller was removing the Calcium ions from my model, even though I put them in the alignment file.  I was able to reproduce the problem using a much smaller file, and I found that I had to do two things to get the model to retain the ions. One was to SET HETATM_IO = on  in the .top file. I guess this was mentioned somewhere in the archives, but it was certainly hard to find.

The other thing that seemed to be necessary was to put the heteroatom listing in the source .pdb file as follows:
HETATM  542  CA  CAL    71     -13.638  53.352  31.500  1.00 12.29      
When I changed the name of the atom from CAL to CA, then it worked. But the file it produced had  converted this back again:
HETATM  538  CAL CAL    71      -9.629  53.605  27.498  1.00 67.12      
I'm not sure what is going on here.

The log file still has a pair of errors:
delete__442E> One or more atoms absent from MODEL:  O:  70: C:  70: N:  71: CA:  71:
delete__442E> One or more atoms absent from MODEL:  O:  71: C:  71: N:  72: CA:  72:
The 2 ions are residue numbers 71 and 72, and it seems to be trying to find amino acid backbones for them!

There are also a number of warnings in the log file, which can be seen by looking in the .log file in the attached zipped archive. It seems to be producing an adequate model now, but if anyone can tell me what is going on, I would appreciate it.
John Penniston