Modular arrangement of proteins as inferred from analysis of homology

• Published in: Protein science 1994-03, Vol.3 (3), p.482-492
• Main Authors: Sonnhammer, Erik L.L., Kahn, Daniel
• Format: Article
• Language: English
• Subjects: Amino Acid Sequence; Bacterial Proteins - chemistry; Bacterial Proteins - genetics; Biological Evolution; Databases, Factual; domain database; domain families; Life Sciences; Molecular Sequence Data; Molecular Structure; Other; Phosphoenolpyruvate Sugar Phosphotransferase System - genetics; protein domains; protein evolution; protein homology; Proteins - chemistry; Proteins - classification; Proteins - genetics; Sequence Alignment; Sequence Homology, Amino Acid; Software; Software Design
• ISSN: 0961-8368; 1469-896X
• DOI: 10.1002/pro.5560030314

The structure of many proteins consists of a combination of discrete modules that have been shuffled during evolution. Such modules can frequently be recognized from the analysis of homology. Here we present a systematic analysis of the modular organization of all sequenced proteins. To achieve this we have developed an automatic method to identify protein domains from sequence comparisons. Homologous domains can then be clustered into consistent families. The method was applied to all 21,098 nonfragment protein sequences in SWISS‐PROT 21.0, which was automatically reorganized into a comprehensive protein domain database, ProDom. We have constructed multiple sequence alignments for each domain family in ProDom, from which consensus sequences were generated. These nonredundant domain consensuses are useful for fast homology searches. Domain organization in ProDom is exemplified for proteins of the phosphoenolpyruvate: sugar phosphotransferase system (PEP:PTS) and for bacterial 2‐component regulators. We provide 2 examples of previously unrecognized domain arrangements discovered with the help of, ProDom.