ABSTRACT
This chapter discusses statistical design's nature and the principles they are based on and considers the important aspects of the experimental design, based on the analytical methods. The data distribution will be far from being Gaussian, a fundamental pre-requisite for the analysis of the microarray data. The missing values have generally two possible origins: the microarray contains a defect resulting in the impossibility of taking a reading or the machine eliminates the measurement as the value is very close to the noise level. A more sophisticated approach is Principal component analysis (PCA), first introduced in 1901 by Pearson. The mixing matrix at the top was calculated by the spreadsheet, the mixing matrix at the bottom by PCA. “Independent component analysis tries to find a linear representation of non-Gaussian data so that the components are statistically independent or as independent as possible”. Microarrays are sometimes seen as the miracle tool, which will give all the answers to all the questions.
