基于熵和Fold Change算法的抗前列腺癌药物重定位分析

Prostate Cancer Drugs Repositioning Based on Entropy and Fold Change Algorithm

摘要: 本文针对抗前列腺癌药物重定位问题，首先将R软件中的信息熵和差异表达倍数Fold Change算法结合起来，分析前列腺癌的基因表达数据，并提取含有大量致病基因信息的前列腺癌特征基因，通过FC算法得到2788条上调基因和2222条下调基因；然后对这些特征基因各筛选出200条具有显著信息的特征基因作为检索标签进行cmap (https://portals.broadinstitute.org/cmap/)分析；最后通过cmap负相关分数降序排列，分析确定几种可能对治疗前列腺癌有效的药物。与传统的提取特征基因方法相比，信息熵和Fold Change算法相结合能减少计算步骤和计算量，成本低、耗时少、准确度高。该方法有效可行。

Abstract: Drug repositioning research not only greatly reduces the cycle of new drug development, but also reduces the economic cost. Prostate cancer is a cancer with high mortality in the world. In recent years, the research on single gene is not enough to analyze the complex pathogenesis of cancer. In this paper, information entropy and Fold Change (FC) algorithm in R software were applied to an-alyze the gene expression data of prostate cancer and extract the characteristic genes of prostate cancer with a large number of pathogenic gene information. Then several drugs for prostate cancer were identified by cmap analysis. The gene expression data used in this paper were from TCGA, and 2788 up-regulated genes and 2222 down-regulated genes were obtained through FC algorithm. And characteristics of these genes screened 200 select genes with significant information as retrieval label cmap (https://portals.broadinstitute.org/cmap/) analysis; finally, the negative correlation score of cmap was ranked in descending order, and several possible effective drugs for the treatment of prostate cancer were analyzed and determined. Compared with the traditional method of extracting characteristic genes, the combination of information entropy and Fold Change algorithm can reduce the computational steps and amount of computation, with low cost, low time consumption and high accuracy. This method is effective and feasible.