The search is independent of any connectivity between atoms, as well as on atom type, name or input order.
The search for correspondances is achieved atom by atom (not residue by residue).
This service will search for the occurrence of Motifs of a QUERY structure in a collection of template structures (the BANK).
The query must be a PDB file.
To specify the BANK, you can either directly input a multiPDB file (HEADER and END fieds are mandatory), or paste PDB ids, URL of PDB files or of lists of PDB ids. PDB Ids one the form 1tim for the complete 1tim entry (includes chains A and B), 1timA for chain A only.
More help here.