[Bank of samples from the Prof_Pat protein family, assessment of efficacy].

Autor: Nizolenko LF; Institute of Molecular Biology, State Research Center of Virology and Biotechnology VECTOR, Russian Ministry of Health, Koltsovo, Novosibirsk Region, 630559 Russia. nizolenko@vector.nsc.ru, Bachinskiĭ AG, Naumochkin AN, Iarygin AA, Grigorovich DA
Jazyk: ruština
Zdroj: Molekuliarnaia biologiia [Mol Biol (Mosk)] 2004 Mar-Apr; Vol. 38 (2), pp. 256-64.
Abstrakt: The PROF_PAT protein pattern database has been created and maintained so as to comprise the maximal number of the SWISS-PROT + TrEMBL proteins as patterns. The present paper describes some characteristic features of PROF_PAT to assist the potential user. New amino acid sequences (10938) from the SWISS-PROT database have been analyzed to determine the boundary values of the "score" parameter to distinguish random and significant similarities. Analysis through the Internet of 20 amino acid sequences having no descriptions in the TrEMBL database demonstrated that PROF_PAT, being highly competitive with its counterparts in specificity, surpasses them in amplitude and variety of proteins, working several times as fast. The real representation of protein families in the PROF_PAT database (release 1.11), which contains 50,149 patterns of 344,429 proteins, has been estimated at 31,450.
Databáze: MEDLINE