I'm working on an algorithm that will determine the conserved regions like sequence level domains for the input protein sequences. I am evaluating my method by comparing the results with the domains annotated in Pfam. However, I am asked to confirm my results by comparing them with PDB. I'm looking into PDB and I don't see a way to extract domain regions from it, are the "biological assemblies" in PDB the same as domains? Is there a way to extract the domain regions from PDB for an input set of sequences (say a fasta file)?