Impact of ontology evolution on functional analyses
Impact of ontology evolution on functional analyses
复制标题
DOI：
10.1093/bioinformatics/bts498
复制
发表时间：
2012-10-15
期刊：
影响因子：
5.8
通讯作者：
Rahm, Erhard
中科院分区：
文献类型：
--
作者：
Gross, Anika;Hartung, Michael;Rahm, Erhard
动机：本体论用于生物数据的注释和分析。随着知识的积累，本体和注释经历不断的修改以反映这些新知识。这些修改可能会影响统计应用的结果，例如用本体论分组描述实验数据的功能富集分析。在这里，我们研究基因本体（GO）的修改在多大程度上影响这些实验和模拟数据的统计分析。该分析基于结果集稳定性的新度量，并考虑了不同的本体和注释变化。结果：我们的研究结果表明，GO过去的变化在本体的不同分支上分布不均匀。考虑分析结果中重要类别的语义相关性，可以对功能富集研究进行更现实的稳定性评估。我们观察到，尽管本体和注释发生了变化，术语富集分析的结果却趋于惊人的稳定。
Motivation: Ontologies are used in the annotation and analysis of biological data. As knowledge accumulates, ontologies and annotation undergo constant modifications to reflect this new knowledge. These modifications may influence the results of statistical applications such as functional enrichment analyses that describe experimental data in terms of ontological groupings. Here, we investigate to what degree modifications of the Gene Ontology (GO) impact these statistical analyses for both experimental and simulated data. The analysis is based on new measures for the stability of result sets and considers different ontology and annotation changes.Results: Our results show that past changes in the GO are non-uniformly distributed over different branches of the ontology. Considering the semantic relatedness of significant categories in analysis results allows a more realistic stability assessment for functional enrichment studies. We observe that the results of term-enrichment analyses tend to be surprisingly stable despite changes in ontology and annotation.