Signal values from Feature Extraction output were imported into an Excel sheet. The selection of the data was based on a total of 44 arrays, which included datasets from previous studies. Control spots were removed; the gMedianSignal and gBGMedianSignal data were used for selecting those spots that showed a signal higher than 2x the background on all 44 arrays (total of 20,370 spots selected). 22 Arrays were used in the present study. Values were Log-2 transformed. Data were normalized per array by subtracting every value by the median value of that array; then data were normalized per spot by subtracting the median value of all 22 arrays.
Slides were scanned after excitation of Cy3 at 543 nm in Agilent C Scanner, using the default settings of the equipment. Scanned Tiff images were imported into Feature Extraction software v8.5 and visually inspected prior to subsequent analysis. A grid template (015425_D_F_20061105) was fiited to the image. The fluorescent intensity and background signal were examined for each array to determine array quality.
cRNAs were randomly distributed over the arrays of 4x44K POCI slides and hybridized for 17h at 65°C in an Agilent Hybridization oven according to One-Color Microarray-Based Gene Expression Analysis (Low Input Quick Amp Labeling) Protocol for use with Agilent Gene Expression oligo microarrays Version 6.5, May 2010.
Purified RNA (25 µg) was labelled by incorporation of Cy3-dCTP during cDNA synthesis using Low Input Quick Amp Labeling kit according to One-Color Microarray-Based Gene Expression Analysis Protocol for use with Agilent Gene Expression oligo microarrays Version 6.5, May 2010.
Five varieties of potato plants (Biogold, Fontane, Innovator, Lady Rosetta and Maris Piper) were grown in Wageningen, the Netherlands (NL) and harvested in 2010. One genotype of potato (Sante) was grown in the United Kingdom (UK) as part of the QLIF study (van Dijk, 2012, J. Agric. Food Chem.: 60, 2090-2101, E-MTAB-605) and harvested in 2005. Each of the NL cultivars was farmed in two lots of different fields; one lot was raised in clay and the other one in sand, having two individual plots per cultivar, and in each plot, four individual plants. The exception was Maris Piper, which in both fields was grown in one single plot containing three individual plants. All varieties were grown under conventional farming conditions. The only exception was Sante_OrgCropProtection_ConvFertilizer_rep3, which was grown under conventional fertilizer and organic crop protection, renamed in in the present study as Sante_UK_2005_OCP_CF_rep3.
Total RNA was isolated from 0.5 g of each freeze-dried sample according to the protocol for isolation of RNA from freeze-dried plant material (van Dijk, 2009, J. Agric. Food Chem.: 57, 1612-1623), based on lysis with a hexadecyltrimethylammonium bromide (CTAB) buffer, followed by chloroform/isoamyl alcohol extraction and overnight precipitation with lithium chloride. The modifications were: the extraction buffer was pre-warmed to 60°C before used, the chloroform/isoamyl alcohol extraction was repeated three times before the LiCl precipitation and the final precipitation with 96% ethanol was performed with the tubes kept on ice and then centrifuged at 4°C for 15 min at 14000 g. Total RNA isolated was dissolved in 100 ?L of 10 mM Tris (pH 7) and warmed to 65°C for 10 min.