Tu Kang, Yu Hui, Guo Zheng, Li Xia
Department of Bioinformatics, Harbin Medical University, Harbin 150086, People's Republic of China.
Genomics. 2004 Dec;84(6):922-8. doi: 10.1016/j.ygeno.2004.08.005.
Currently the functional annotations of many genes are not specific enough, limiting their further application in biology and medicine. It is necessary to push the gene functional annotations deeper in Gene Ontology (GO), or to predict further annotated genes with more specific GO terms. A framework of learnability-based further prediction of gene functions in GO is proposed in this paper. Local classifiers are constructed in local classification spaces rooted at qualified parent nodes in GO, and their classification performances are evaluated with the averaged Tanimoto index (ATI). Classification spaces with higher ATIs are selected out, and genes annotated only to the parent classes are predicted to child classes. Through learnability-based further predicting, the functional annotations of annotated genes are made more specific. Experiments on the fibroblast serum response dataset reported further functional predictions for several human genes and also gave interesting clues to the varied learnability between classes of different GO ontologies, different levels, and different numbers of child classes.
目前，许多基因的功能注释不够具体，限制了它们在生物学和医学中的进一步应用。有必要在基因本体论（GO）中进一步深入基因功能注释，或者用更具体的GO术语预测更多注释基因。本文提出了一种基于可学习性的GO中基因功能进一步预测框架。在以GO中合格父节点为根的局部分类空间中构建局部分类器，并用平均谷本指数（ATI）评估其分类性能。选择ATI较高的分类空间，并将仅注释到父类别的基因预测到子类。通过基于可学习性的进一步预测，使已注释基因的功能注释更加具体。对成纤维细胞血清反应数据集的实验报告了对几个人类基因的进一步功能预测，也为不同GO本体、不同层次和不同子类数量的类别之间不同的可学习性提供了有趣的线索。
