Mouse Exons

Exon tiling array encompassing 837,251 probes across the mouse genome. 12 tissue pools. Composition of the 12 mRNA pools analyzed (mRNA per array hybridization): 1- Heart (2 microgram), Skeletal muscle (2 microgram) 2- Liver (2 microgram) 3- Whole brain (1.5 microgram), Cerebellum (0.48 microgram), Olfactory bulb (0.15 microgram) 4- Colon (0.96 microgram), Intestine (1.04 microgram) 5- Testis (3 microgram), Epididymis (0.4 microgram) 6- Femur (0.9 microgram), Knee (0.4 microgram), Calvaria (0.06 microgram), Teeth+mandible (1.3 microgram), Teeth (0.4 microgram) 7- 15 day Embryo (1.3 microgram), 12.5 day Embryo (12.5 microgram), 9.5 day Embryo (0.3 microgram), 14.5 day Embryo head (0.25 microgram), ES cells (0.24 microgram) 8- Digit (1.3 microgram), Tongue (0.6 microgram), Trachea (0.15 microgram) 9- Pancreas (1 microgram), Mammary gland (0.9 microgram), Adrenal gland (0.25 microgram), Prostate gland (0.25 microgram) 10- Salivary gland (1.26 microgram), Lymph node (0.74 microgram) 11- 12.5 day Placenta (1.15 microgram), 9.5 day Placenta (0.5 microgram), 15 day Placenta (0.35 microgram) 12- Lung (1 microgram), Kidney (1 microgram), Adipose (1 microgram), Bladder (0.05 microgram) Labeling and hybridization: mRNA pools were reverse-transcribed with random nonamer primers (1 microgram/reaction) and T18VN (0.25 microgram/reaction) to synthesize cDNA. The reaction contained a 1:1 mixture of 5-(3-aminoallyl) thymidine 5'-triphosphate (Sigma, St. Louis, MO) and thymidine triphosphate (TTP) in place of TTP. cDNA products were bound to QIAquick PCR Purification columns (Qiagen, Germany) following the manufacturer's instructions, washed three times with 80% ethanol, and eluted with water. Purified cDNA was reacted with N-hydroxy succinimide esters of Cy3 or Cy5 (Amersham Pharmacia Biotech, Piscataway, NJ) following the manufacturer's instructions. Hydroxylamine-quenched Cy-labeled cDNAs were separated from free dye molecules using QIAquick columns. Mixed labeled cDNAs were added to hybridization buffer containing 1 M NaCl, 0.5% sodium sarcosine, 50 mM methyl ethane sulfonate (MES), pH 6.5, 33% formamide and 40 µg herring sperm DNA (Invitrogen, Carlsbad, California). Hybridizations were carried out in a final volume of 0.5 ml injecting into an Agilent hybridization chamber at 42°C on a rotating platform in a hybridization oven (Robbins Scientific Corporation, Sunnyvale, California) for 16-24 h. Slides were then washed (rocking ~30 seconds in 6x SSPE, 0.005% sarcosine, then rocking ~30 seconds in 0.06x SSPE) and scanned with a 4000A microarray scanner (Axon Instruments, Union City, California). Each array was hybridized with two samples simultaneously, each from a different pool. Image processing and data normalization: TIFF images were quantitated with GenePix (Axon Instruments). Individual channels were spatially detrended (i.e overall correlations were removed between spot intensity and position on the slide) by subtracting from each spot's log intensity a weighted average of the log intensities of nearby spots. Most weights were a product of two Gaussian functions, one of the difference in rows between the spots and one of the difference in columns, however 10% of the spots ("high expressors") were always assigned a weight of zero. The widths of the row and column Gaussians were chosen to minimize the sum, across all spots, of the squared difference between the weighted average and the actual log intensity of the spots. The "high expressors" spots were those with the largest difference between their log intensity and the median log intensity over a 7x7 window of neighboring spots. After spatial detrending, we transformed the log intensities back into linear intensities and all single-channels were normalized to each other (in the linear domain), using channel-specific affine transforms. That is, for each single channel, we normalized all the detrending intensities by adding an offset parameter (which was possibly negative) and then multiplying by a positive scale parameter. The scale and offset parameters of each channel's transform were set so that the median intensities and the difference between the 75th and 25th percentile intensities were identical across channels.