I am using Phaser MR for solving a homodimer DNA binding protein with dsDNA complex structure. I'd like to use sequence file rather then molecular weight in the component definition of ASU.  As there are solved homodimer protein structures (amino acid identity 100%) on PDB, I can download seq file (fasta form) from PDB webpage, but the seq file is only for one protein chain.  I also tried online tools like "make sequence from PDB file" (http://swift.cmbi.ru.nl/servers/html/soupir.html), when I put the protein dimer bound to dsDNA pdb into the online tool, the outtput pir sequence file only shows a single protein chain and dna which dose not seems like two strands, rather then just put two 23base dna together for a 46 base length single strand DNA.

I am asking is there a way to generate sequence file for a homodimer and dsDNA? or I have to generate sequence file for each chain separately (there will be four sequence files in this case, two protein chains and two dna chains)?