Abstract: The invention relates generally to biotechnology, and more specifically to in silico methods of identifying lead molecules that have an increased probability of becoming an approved medicament, and business methods of identifying molecules such that they have an increased probability of becoming an approved medicament. Provided is a method for identifying a biologically active peptide consisting of two to seven amino acid residues, comprising the steps of providing a database comprising a plurality of naturally occurring polypeptide sequences; defining at least one peptide motif that satisfies certain defined criteria; and determining for the defined peptide motif its frequency of occurrence among the polypeptide sequences in the database and correlating the frequency with the biological activity of the peptide.