2.5. Data Extraction, Normalisation and Comparisons
Raw microarray data were extracted from scanned images by using the Feature Extraction software (version 10.7.31; Agilent) [26]. Background normalization was carried out using the normexp algorithm whereas normalization between the samples was carried out using quantile normalization [27,28]. Both normalization steps were implemented using the R package of limma (Linear Models for Microarray Data, version 2.13.1) [29]. The remaining parameters in limma were set to default value. For an expression difference to be determined as statistically significant, the log fold change must be ≥0.6 and false discovery rate (FDR) ≤ 0.05. Comparisons of selected candidates between oil palm mesocarp, rice and the Arabidopsis microarray were carried out based on the normalized signal intensities produced.