Kadanga Ali Kpatcha, Leroux Christine, Bonnet Muriel, Chauvet Stéphanie, Meunier Bruno, Cassar-Malek Isabelle, Hocquette Jean-François
INRA, UR1213, Unité de Recherches sur les Herbivores, Centre de Recherches de Clermont-Ferrand/Theix, F-63122 Saint Genès-Champanelle, France.
Gene Regul Syst Bio. 2008 Mar 17;2:107-12. doi: 10.4137/grsb.s414.
This study was conducted with the aim of optimizing the experimental design of array experiments. We compared two image analysis and normalization procedures prior to data analysis using two experimental designs. For this, RNA samples from Charolais steers Longissimus thoracis muscle and subcutaneous adipose tissues were labeled and hybridized to a bovine 8,400 oligochip either in triplicate or in a dye-swap design. Image analysis and normalization were processed by either GenePix/MadScan or ImaGene/GeneSight. Statistical data analysis was then run using either the SAM method or a Student's t-test using a multiple test correction run on R 2.1 software. Our results show that image analysis and normalization procedure had an impact whereas the statistical methods much less influenced the outcome of differentially expressed genes. Image analysis and data normalization are thus an important aspect of microarray experiments, having a potentially significant impact on downstream analyses such as the identification of differentially expressed genes. This study provides indications on the choice of raw data preprocessing in microarray technology.
本研究旨在优化阵列实验的实验设计。在数据分析之前，我们使用两种实验设计比较了两种图像分析和标准化程序。为此，将夏洛来牛胸最长肌和皮下脂肪组织的RNA样本进行标记，并以一式三份或染料交换设计与牛8400寡核苷酸芯片杂交。图像分析和标准化由GenePix/MadScan或ImaGene/GeneSight进行处理。然后使用SAM方法或使用R 2.1软件上运行的多重检验校正的学生t检验进行统计数据分析。我们的结果表明，图像分析和标准化程序有影响，而统计方法对差异表达基因的结果影响较小。因此，图像分析和数据标准化是微阵列实验的一个重要方面，对下游分析如差异表达基因的鉴定可能有重大影响。本研究为微阵列技术中原始数据预处理的选择提供了指导。
