Evaluation of Function Predictions for Moonlighting Proteins

<p>The performances of PFP, ESG, and PSI-BLAST in predicting the functional diversity of moonlighting proteins were analyzed. PFP shows overall better performance in predicting diverse moonlighting functions as compared with PSI-BLAST and ESG. Recall by PSI-BLAST greatly improved when BLOSUM45 was used. This analysis indicates that considering weakly similar sequences in prediction enhances the performance of sequence based AFP methods in predicting functional diversity of moonlighting proteins. The current study will also motivate development of novel computational frameworks for automatic identification of such proteins. This dataset was used for the evaluation of function predictions for moonlighting proteins. The dataset provides a set of 19 moonlighting proteins from Huberts et al. (2010) with GO annotations taken for each protein from Uniprot. The GO annotations for the proteins are classified as in “Function 1”, “Function 2”, “Function Common to Both” and “Function Not Clear”. The dataset is available in two excel files: the first file contains the protein names and uniprot ID, and the second file contains the current GO annotations for the proteins from Uniprot, functional description, and functional classification of the GO terms.</p>

本研究分析了PFP、ESG及位置特异性迭代BLAST（PSI-BLAST）在预测兼性多功能蛋白（moonlighting proteins）功能多样性方面的表现。相较于PSI-BLAST与ESG，PFP在预测多样化兼性多功能蛋白功能方面整体表现更优。当使用BLOSUM45置换矩阵时，PSI-BLAST的召回率（Recall）得到大幅提升。本分析表明，在预测流程中纳入弱相似序列，可提升基于序列的自动功能注释（Automatic Function Prediction，AFP）方法对兼性多功能蛋白功能多样性的预测性能。本研究亦将推动用于自动识别此类蛋白的新型计算框架的开发。本数据集用于评估兼性多功能蛋白的功能预测效果。该数据集涵盖来自Huberts等人2010年研究的19种兼性多功能蛋白，各蛋白的基因本体（Gene Ontology）注释信息均从UniProt数据库获取。上述蛋白的基因本体注释被划分为"功能1""功能2""二者共有功能"及"功能未明确"四类。本数据集以两个Excel文件形式提供：第一个文件包含蛋白名称与UniProt ID，第二个文件收录了各蛋白从UniProt获取的最新基因本体注释、功能描述及基因本体术语的功能分类。