Hiller Karsten, Grote Andreas, Scheer Maurice, Münch Richard, Jahn Dieter
Institut für Mikrobiologie, Technische Universität Braunschweig, Spielmannstrasse 7, D-38106 Braunschweig, Germany.
Nucleic Acids Res. 2004 Jul 1;32(Web Server issue):W375-9. doi: 10.1093/nar/gkh378.
We have developed PrediSi (Prediction of Signal peptides), a new tool for predicting signal peptide sequences and their cleavage positions in bacterial and eukaryotic amino acid sequences. In contrast to previous prediction tools, our new software is especially useful for the analysis of large datasets in real time with high accuracy. PrediSi allows the evaluation of whole proteome datasets, which are currently accumulating as a result of numerous genome projects and proteomics experiments. The method employed is based on a position weight matrix approach improved by a frequency correction which takes in to consideration the amino acid bias present in proteins. The software was trained using sequences extracted from the most recent version of the SwissProt database. PrediSi is accessible via a web interface. An extra Java package was designed for the integration of PrediSi into other software projects. The tool is freely available on the World Wide Web at http://www.predisi.de.
我们开发了PrediSi(信号肽预测工具),这是一种用于预测细菌和真核生物氨基酸序列中信号肽序列及其切割位置的新工具。与之前的预测工具不同,我们的新软件对于实时高精度分析大型数据集特别有用。PrediSi可对目前因众多基因组计划和蛋白质组学实验而不断积累的全蛋白质组数据集进行评估。所采用的方法基于一种通过频率校正改进的位置权重矩阵方法,该方法考虑了蛋白质中存在的氨基酸偏差。该软件使用从最新版SwissProt数据库中提取的序列进行训练。可通过网络界面访问PrediSi。还设计了一个额外的Java包,用于将PrediSi集成到其他软件项目中。该工具可在万维网的http://www.predisi.de上免费获取。