Protein Identification By Combining MS With Database Search.
 Mass spectrometry has been combined with database search to create a valuable and automatic protein identification tool. Many three types of databases (DNA, cDNA/EST and nrdb) are searched by mass spectrometric data.

Non-redundant protein database (nrdb) contains known sets of full length protein sequences, devoid of duplicates. In this application, intact proteins are degraded into pools of peptides whose masses are determined by mass spectrometry and then searched against genomic DNA, cDNA/EST and nrdb database entries. These databases also be searched both by mass fingerprints and tandem mass spectrometric data.

Protein modifications do not present an obstacle to identification because, for a typical 50 kD protein generally 50 peptides are obtained after tyrpic digestion and any a few are modified. Only a small number of peptides are actually required for unique matching to a database entry, especially in case of data from tandem mass spectrometery; therefore very extensive modification only marginally increases the difficulty of protein identification.Three main programs that are used for database mining involving protein identification using MS data are listed.

Three main programs for protein identification using database mining.

Program	Web site	Based on
Moscot	htt:/?www.matrixscience.com	It incorporates a probability based implementation of the MOWSE* algorithm.
ProFound	http://prowl.rockefeller.edu	It calculates the probability that a protein in a database is the protein analysed here and it uses Z-score as an indicator of the quality of the search result.
Protein Prospector (MS-Fit)	htt://prospector.ucsf.edu	It uses MOWSE* score to evaluate a hit in protein identification.

*The Molecular Weigh SEarch algorithm (MOWSE) capitalizes on information obtained through mass spectrometry (MS) technique, using molecular weights of both, the intact proteins and those resulting from digestion of the same proteins with specific protease).