The PMD is based on the literature, not on proteins. That is, each entry in the database corresponds to one article, which contains several or a number of protein mutants. Each database entry is identified by a serial number and is defined as either natur al or artificial, depending on the type of the mutation. For each entry the following ite ms are recorded : "JOURNAL", "TITLE", "CROSS-REFERENCE", "PROTEIN", "N-TERMINAL", "CHANGE", "FUNCTION", "STRUCTURE", "STABILITY", etc. "CROSS-REFERENCE" indicates the code names of the protein given in other databases such as Protein Identification Resources (2). "N-TERMINAL" shows the N-terminal sequence of five amino acids which may help to show the unambiguous numbering of th e sequence. "CHANGE" indicates the position and kind of mutations, such as amino acid substitut ion, insertion and deletion, denoted with a specific notation. Any functional or structural features ("FUNCTION", "STRUCTURE", "STABILITY",etc) ob served in the mutant are described immediately after 'CHANGE'. Relative differences in activity and/or stability, in comparison with the wild-type protein, are indicated with symbols [- -],[-],[=],[+] or [+ +]. Complete loss of activity is denoted as .
T.Kawabata, M.Ota, and K.Nishikawa. "The protein mutant database"(1999). Nucleic Acids Research ,vol. 27, 355-357.