Rapid identification of proteins.

The amino acid composition, molecular weight, and isoelectric point of a protein can all be easily and economically determined by current electrophoretic techniques. A method which uses such easily obtained data to identify proteins is described. A computer program first corrects for systematic errors in amino acid quantitation and then searches the current sequence database for proteins with amino acid compositions similar to the corrected values, taking into account the reliability of determination of each amino acid. The program also provides the calculated molecular weight, isoelectric point, and name of each candidate, providing three further independent criteria for protein identification. The program is surprisingly sensitive, and the composition data alone, if of good quality, usually suggest the correct protein as a strong candidate if it or a close homologue is present in the database. Further studies show that proteins in the current database have amino acid compositions distinct enough to allow this method to be generally applicable. The method is a quick and cost-effective first step in protein characterization and should become increasingly useful as the number of fully sequenced proteins continues to rise.