乳腺癌基因药物网络模型的构建与分析

Construction and analysis of a breast cancer gene-drug network model

目的构建乳腺癌基因药物网络模型,提取并预测乳腺癌相关基因药物间的关联。方法基于"ABC理论"和关联规则,提出一种生物实体间关联算法,以乳腺癌为例,提取乳腺癌相关基因与基因、药物与药物、基因与药物3种不同层次的关联,采用R语言实现网络模型的可视化,最后利用ROC曲线验证算法可靠性。结果得到乳腺癌相关基因185种,98种不同关联;乳腺癌相关药物97种,170种不同关联;乳腺癌相关基因与药物网络中含有127种基因和77种药物,共384种不同关联。结论乳腺癌的基因药物之间存在大量不同强度的关联,并且发现一些具有高度关联但尚未验证的生物实体对,为乳腺癌个性化诊治提供了新的研究思路。

Objective To construct a breast cancer gene- drug network model for extracting and predicting the correlations between breast cancer- related genes and drugs. Methods We developed an algorithm based on the ABC principle and the association rules to obtain the correlations between the biological entities. For breast cancer, we constructed 3 different correlations（gene-gene, drug-drug and gene-drug） and used the R language to implement the associated network model. The reliability of the algorithm was verified by ROC curve. Results We identified 185 breast cancer- associated genes and 98 associations between them, 97 drugs and 170 associations between them. The breast cancer genes-drugs network contained 127 genes and 77 drugs with 384 associations between them. Conclusion We identified a large number of different correlations between the breast cancer- related genes and drugs and close correlations between some biological entity pairs that have not yet been reported, which may provide a new strategy for experimental design for testing personalized breast cancer treatment.