Using this tool, you can iteratively pile up PDB formatted files into one single multiPDB file, with entries separated by HEADER and END lines. You can specify:
explicit PDB files (upload or paste).
PDB ids (ex: 1tim, 1timA for chain A of 1tim)
URLs to PDB files.
Several filters can apply:
amino acids: you can select only the backbone or side chain atoms. The default is to consider all atoms.
Heteros: you can choose to remove none, all of them or only the solvent.
Accessibility: you can select (on a per residue basis) external or buried residues (according to the relative accessibility threshold specified). This filter can be mixed with the atomic filters given that exposure filter only applies to amino-acids (i.e. heteros but solvent are taken into account for exposure calculation, but are unaffected by this filter).
The files will be piled up in the pending PDB, latest files at the end of the pending PDB. A tool to setup lists of PDB identifiers matching the PDBsum keywords (such as "cysteine protease" can be accessed here