对人类基因表达变异性的大规模分析表明，高度可变的药物靶点与较低的药物有效性和安全性相关。
Large-scale analysis of human gene expression variability associates highly variable drug targets with lower drug effectiveness and safety.
作者信息
Simonovsky Eyal, Schuster Ronen, Yeger-Lotem Esti
机构信息
Department of Clinical Biochemistry & Pharmacology, Ben-Gurion University of the Negev, Beer-Sheva, Israel.
National Institute for Biotechnology in the Negev, Ben-Gurion University of the Negev, Beer-Sheva, Israel.
出版信息
Bioinformatics. 2019 Sep 1;35(17):3028-3037. doi: 10.1093/bioinformatics/btz023.
MOTIVATION
The effectiveness of drugs tends to vary between patients. One of the well-known reasons for this phenomenon is genetic polymorphisms in drug target genes among patients. Here, we propose that differences in expression levels of drug target genes across individuals can also contribute to this phenomenon.
RESULTS
To explore this hypothesis, we analyzed the expression variability of protein-coding genes, and particularly drug target genes, across individuals. For this, we developed a novel variability measure, termed local coefficient of variation (LCV), which ranks the expression variability of each gene relative to genes with similar expression levels. Unlike commonly used methods, LCV neutralizes expression levels biases without imposing any distribution over the variation and is robust to data incompleteness. Application of LCV to RNA-sequencing profiles of 19 human tissues and to target genes of 1076 approved drugs revealed that drug target genes were significantly more variable than protein-coding genes. Analysis of 113 drugs with available effectiveness scores showed that drugs targeting highly variable genes tended to be less effective in the population. Furthermore, comparison of approved drugs to drugs that were withdrawn from the market showed that withdrawn drugs targeted significantly more variable genes than approved drugs. Last, upon analyzing gender differences we found that the variability of drug target genes was similar between men and women. Altogether, our results suggest that expression variability of drug target genes could contribute to the variable responsiveness and effectiveness of drugs, and is worth considering during drug treatment and development.
AVAILABILITY AND IMPLEMENTATION
LCV is available as a python script in GitHub (https://github.com/eyalsim/LCV).
SUPPLEMENTARY INFORMATION
Supplementary data are available at Bioinformatics online.
动机
药物的有效性在患者之间往往存在差异。造成这种现象的一个众所周知的原因是患者之间药物靶基因的遗传多态性。在此，我们提出个体间药物靶基因表达水平的差异也可能导致这种现象。
结果
为了探究这一假设，我们分析了个体间蛋白质编码基因，特别是药物靶基因的表达变异性。为此，我们开发了一种新的变异性度量方法，称为局部变异系数（LCV），它根据每个基因相对于具有相似表达水平的基因的表达变异性进行排名。与常用方法不同，LCV在不强制对变异进行任何分布的情况下消除了表达水平偏差，并且对数据不完整性具有鲁棒性。将LCV应用于19种人体组织的RNA测序图谱和1076种获批药物的靶基因，结果显示药物靶基因的变异性显著高于蛋白质编码基因。对113种具有可用有效性评分的药物进行分析表明，靶向高变异性基因的药物在人群中的效果往往较差。此外，将获批药物与已退出市场的药物进行比较发现，已退出市场的药物所靶向的基因变异性显著高于获批药物。最后，在分析性别差异时，我们发现男性和女性药物靶基因的变异性相似。总之，我们的结果表明，药物靶基因的表达变异性可能导致药物反应性和有效性的差异，在药物治疗和开发过程中值得考虑。
可用性和实现方式
LCV可作为一个Python脚本在GitHub上获取（https://github.com/eyalsim/LCV）。
补充信息
补充数据可在《生物信息学》在线获取。
Zotero 插件