This database has the positional information in the amino acid motif structure. The information is identified by experiments or sequence analysis (homology search and multiple alignment). We use PROSITE and BLOCKS as amino acid motif databases. PROSITE has motives from experiments, and BLOCKS has motives from both experiments and sequence analysis. We use PDB as the protein structure database and decide the motif position from SEQRES and ATOM sequences. The reason we need SEQRES and ATOM sequences is because those are not matched in some PDB entries.