近年来，随着人类基因组测序的完成、蛋白质组等生物分子学研究的不断深入，国际上许多权威的生物医学数据库也已建成并在不断更新中。这些对外开放的数据信息，为构建各类生物分子网络并进行分析提供了机会，生物信息学和网络医学随之兴起，利用网络分析进行潜在致病基因探寻、药物靶标预测等，成为医药学研究和开发的新思路。 本文以癌症疾病基因作为研究对象，利用多个数据源的信息进行了数据整合和网络构建，并进行了简单的癌症药物靶标预测。主要研究内容如下： 1. 调研了多个权威生物医学数据库，选取其中关于蛋白质、基因、药物和癌症的数据集并加以整合，得到药物-靶标网络和癌症-基因网络； 2. 在多源数据和蛋白质互作网络的基础上，构建了癌症-药物互作网络并实现了可视化； 3. 利用经典的基于图的随机游走算法，在该互作网络中将蛋白质与确知癌症疾病蛋白的相关度进行打分排序，通过计算得到了潜在癌症疾病蛋白，完成了潜在的癌症药物靶标的预测； 4. 分析发现预测结果中的大部分蛋白质功能与癌症相关，证明了该方法的合理性与有效性，并针对该结果进一步改进。

With the finish of the human genome sequencing and the great progress in molecular biology like proteomics, many established authoritative international biomedical databases are completing continually in recent years. The opening of these databases gives a chance for building all kinds of biological molecular networks, which promotes the development of information biology and network medicine. Applying network-based approaches in potential disease gene detection and drug target prediction provides a new idea for medical research and drug production. This paper integrates datasets from several different databases and establishes networks focusing on cancer genes, and makes a simple cancer drug target prediction further. The main contribution of this paper are listed below: 1. Several authoritative biomedical databases are introduced firstly in this study. Datasets about proteins, genes, drugs and cancers are integrated as multi-source data to build Drug-Target network and Cancer-Gene network. 2. Cancer-Drug Interaction network based on multi-source data and Protein-Protein Interaction network are constructed and visualized. 3. In Cancer-Drug Interaction network we build, affinities between products of known cancer genes and other proteins are scored by using method of Random Walks on graphs. Candidate cancer-related proteins which has high scores are predicted as potential cancer drug targets. 4. The predicting targets are analyzed to verify rationality and efficiency of the method. The analysis shows that most of the predicting targets are associated to cancers or tumours in functions. Improvement measures are proposed at the end.