多实验平台下基因表达数据分析研究

发布时间：2018-10-05 11:00

【摘要】：基因表达分析是转录组学最基本的研究手段之一,对基因和异构体表达水平的计算及差异表达分析,有助于人们了解基因和剪切异构体的功能以及调控机制。作为当前主流的两种大规模基因表达测量技术,基因芯片和基于高通量测序技术的RNA-Seq方法广泛应用于转录组学研究领域,并且产生了海量的表达数据,为多平台表达数据融合提供了可行性。本文的工作主要从以下两方面展开研究:(1)多平台下基因和异构体表达分析对比研究。首先介绍了广泛使用的Affymetrix传统3’基因芯片、外显子芯片、较新的全转录组芯片,以及基于RNA-Seq技术的Illumina平台这四个主流实验平台的技术原理。其次从基因表达水平计算和差异表达分析两方面介绍了每个平台下一些主流数据分析方法,分析了每个平台下各数据分析方法的优劣,并通过标准数据集对比分析了一些代表性方法的性能,获得的对比研究结果为研究者选择实验平台以及表达数据分析方法提供了参考。(2)融合多平台表达数据的转录组差异表达分析。针对现有的多平台差异表达分析研究方法存在的问题,本文提出了融合多平台表达数据的差异表达检测模型mpDE(multi-platform Differential Expression model)。该模型将不同实验平台表达数据和技术性测量误差融入模型中,同时考虑了同一平台在不同条件下的生物重复或技术重复的波动性,从而提高差异表达分析的准确度。本文将mpDE应用到三个人类数据集,并与单平台的差异表达检测结果和其他多平台表达数据融合方法进行了对比。实验结果表明,mpDE能够获得更加准确灵敏的差异表达分析结果。
[Abstract]:Gene expression analysis is one of the most basic research methods in transcriptome. The calculation of gene and isomer expression level and differential expression analysis are helpful to understand the function and regulation mechanism of gene and shear isomer. As two kinds of large-scale gene expression measurement techniques, gene chip and RNA-Seq method based on high-throughput sequencing technology are widely used in the field of transcriptome research, and produce a large amount of expression data. It provides the feasibility for multi-platform expression data fusion. The main works of this paper are as follows: (1) the comparative analysis of gene and isomer expression in multi-platform. Firstly, the technical principles of traditional Affymetrix 3'gene chip, exon chip, new full transcriptome chip and Illumina platform based on RNA-Seq technology are introduced. Secondly, from two aspects of gene expression level calculation and differential expression analysis, this paper introduces some mainstream data analysis methods under each platform, and analyzes the advantages and disadvantages of each data analysis method under each platform. The performance of some representative methods is compared with the standard data set. The results provided a reference for the researchers to choose the experimental platform and the method of expression data analysis. (2) the transcriptome differential expression analysis combined with multi-platform expression data. Aiming at the problems existing in the existing research methods of multi-platform differential expression analysis, a differential expression detection model (mpDE) (multi-platform Differential Expression model).) based on fusion of multi-platform expression data is proposed in this paper. The model integrates different experimental platform data and technical measurement error into the model and considers the volatility of biological repetition or technical repetition of the same platform under different conditions so as to improve the accuracy of differential expression analysis. In this paper, mpDE is applied to three human datasets, and the results of differential expression detection on single platform and other multi-platform expression data fusion methods are compared. The experimental results show that MMP DE can obtain more accurate and sensitive differential expression analysis results.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：Q811.4

【参考文献】