Arrow plot and correspondence analysis maps for visualizing the effects of background correction and normalization methods on microarray data

• Published in: Pattern Recognition in Computational Molecular Biology pp. 394 - 416
• Main Authors: Silva, Carina, Freitas, Adelaide, Roque, Sara, Sousa, Lisete
• Format: Book Chapter
• Language: English
• Published: Hoboken, NJ, USA Wiley 2016; John Wiley & Sons, Inc
• Subjects: Arrow plot; Background correction methods; Correspondence analysis; Differentially expressed genes; Microarray data; Normalization methods; Significance analysis of microarrays
• ISBN: 9781118893685; 1118893689
• DOI: 10.1002/9781119078845.ch21

Among various available array technologies, double-channel cDNA microarray experiments provide numerous technical protocols associated with functional genomic studies. The chapter begins by detailing the arrow plot, which is a recent graphical-based methodology to detect differentially expressed (DE) genes, and briefly mentions the significance analysis of microarrays (SAM) procedure, which is, in contrast, quite well known. Next, it introduces the correspondence analysis (CA) and explains how the resultant graphic can be interpreted. Then, CA in both class comparison and class prediction applications and over the data sets lymphoma (lym), lung (lun), and liver (liv) is executed. The CA is applied to all three databases in order to obtain graphical representations of background correction (BC) and normalization (NM) profiles in a two-dimensional reduced space. Whenever possible, more than one preprocessing strategy on microarray data could be applied and results from preprocessed data should be compared before any conclusion and subsequent analysis.