| 研究概要 |
In this study, a novel classification scheme was defined for intrinsically disordered regions using their chemical composition. It was shown that the functions of proteins containing intrinsically disordered regions are related, at least in part, with the chemical nature of the disordered regions.
Since the chemical composition of disordered regions by itself was found to be insufficient as a predictor of protein function, several possible sequence features of disordered regions were identified. These sequence features included information about the amino acid composition, chemical composition, amino acid pair frequencies and features associated with predicted secondary structure of the intrinsically disordered region. These sequence features were extracted from disordered regions within proteins and were rigorously tested for their ability to predict protein function.
The study showed that sequence features from intrinsically disordered regions can be used to predict protein function with a combination of features performing better than single ones. Additionally, different combinations of features perform best for the prediction of different functional terms. Further investigation needs to be performed to identify the best combinations of sequence features from disordered regions that are good predictors of protein function. The study also highlighted the inadequate nature of the current functional annotation schemes, specifically Gene Ontology terms, and their inability to capture the various functions of intrinsically disordered regions.
|
