Subcellular localization prediction of proteins containing fibronectin domains using collocation of amino acid pairs / 中国组织工程研究

ABSTRACT

BACKGROUND: Proteins containing fibronectin domains play an important role in cell migration, adhesion, growth and differentiation and have been widely applied to a variety of new biological materials. Subcellular localization prediction of proteins containing fibronectin domains can promote protein function research and development of new biomaterials.OBJECTIVE: To realize subcellular localization prediction of proteins containing fibronectin domains. METHODS: A total of 80 human proteins were randomly selected from Uniprot database. The amino acid pairs for each protein were collocated to form 400 dimensional input feature vectors. The feature vectors were then trained and tested using support vector machine and k-nearest neighbor separately. The prediction quality was examined by the jackknife test. RESULTS AND CONCLUSION: The prediction accuracy was 92.5% and 95% for support vector machine and k-nearest neighbor methods respectively. This suggests that support vector machine and k-nearest neighbor methods are of important significance for predicting subcellular localization of proteins containing fibronectin domains and contribute to functional research of such proteins and surface modification of new biomaterials.