Due to technological advances the quality and availability of biological data has increased dramatically in the last decade. Analysing protein-protein interaction networks (PPINs) in an integrated way, together with subcellular compartment data, provides such biological context, helps to fill in the gaps between a single type of biological data and genes causing diseases and can identify novel genes related to disease. In this study, we present BCCGD, a method for integrating subcellular localization data with PPINs that detects breast cancer candidate genes in protein complexes. We achieve this by defining the significance of the compartment, constructing edge-weighted PPINs, finding protein complexes with a non-negative matrix factorization approach, generating disease-specific networks based on the known disease genes, prioritizing disease candidate genes with a WDC method. As a case study, we investigate the breast cancer but the techniques described here are applicable to other disorders. For the top genes scored by BCCGD approach, we utilize the literature retrieving method to test the correlations of them with the breast cancer. The results show that BCCGD discover some novel breast cancer candidate genes which are valuable references for the biomedical scientists.

中文翻译：

---

通过将亚细胞定位数据与蛋白质-蛋白质相互作用网络相集成来检测乳腺癌候选基因。

由于技术的进步，生物数据的质量和可用性在过去十年中得到了极大的提高。以综合的方式分析蛋白质-蛋白质相互作用网络（PPIN），以及亚细胞区室数据，提供了这种生物学背景，有助于填补单一类型的生物学数据与引起疾病的基因之间的空白，并可以识别与疾病相关的新基因。在这项研究中，我们提出了BCCGD，一种将亚细胞定位数据与PPIN整合的方法，该方法可检测蛋白复合物中的乳腺癌候选基因。我们通过定义区室的重要性，构建边缘加权的PPIN，使用非负矩阵分解方法找到蛋白质复合物，基于已知疾病基因生成特定疾病的网络，使用WDC方法对疾病候选基因进行优先排序。作为案例研究，我们调查了乳腺癌，但此处描述的技术适用于其他疾病。对于通过BCCGD方法得分最高的基因，我们利用文献检索方法来检验它们与乳腺癌的相关性。结果表明，BCCGD发现了一些新的乳腺癌候选基因，为生物医学科学家提供了有价值的参考。