SIMAP is a public database of pre-calculated protein similarities that plays a key role in many bioinformatics methods. It contains about all currently published protein sequences and is continuously updated.