malign3D question

I was trying to use the malign3D command to do an alignment of three sequences based on their PDB structures. I was able to get an inital alignment from the command malign with no problems. I used this alignment as the starting point. I then copied the commands found on the user manual page for malign3d (of course changing the file names as appropriate) to a .top file. When I look at the .log file I get the following error

fndatmi_284E> Beginning atom index for a residue out of bounds: 0 0

I have also tried running align3d for two sequences at a time and that works fine. Any suggestions on what I am doing wrong?

Thanks, Karen Ellis