一种基于基因拓扑重要性的通路识别方法

癌症相关通路的识别是认识癌症发生发展过程机制的生物学基础。已有的通路识别方法很少考虑基因在通路中的拓扑重要性。重叠基因降权(PADOG)方法在基因集分析(GSA)方法的基础上融入了基因特异性的影响,提高了癌症相关通路的识别性能。为进一步提高癌症相关通路的识别性能,首先统计了KEGG通路数据集中基因出度的分布情况,根据基因出度的大小定义了基因的重要性。最后将基因的特异性和重要性融合在一起,提出了一种基于基因重要性和特异性的通路分析方法 PAGIS。在结肠癌、肺癌和胰腺癌3个数据集上的实验结果表明,PAGIS方法比PADOG能够提高很多癌症相关的排名,从而提高癌症相关通路的识别效果。

A pathway analysis method based on the topological importance of genes

Identifying cancer-related pathways is important for understanding the underling mechanisms of the development of cancers. However, current pathway analysis methods are lack of the consideration of topology characteristics of genes in the pathways. Pathway Analysis with Down-weighting of Overlapping Genes (PADOG) considered the specificity of genes based on the GSA method to improve its performance. In order to improve the performance of identifying cancer-related pathways, we first studied the out-degree distribution of genes in the KEGG pathway database. Then we defined the importance of genes based on their out-degree. Finally we proposed a pathway analysis based on the important and specificity of genes (PAGIS). Analysis results from the colorectal cancer datasets showed that our improved method could identify more cancer related pathways than PADOG.