|
| Status |
Public on May 03, 2009 |
| Title |
ago3t2/ago3t3 (replicated 2) |
| Sample type |
RNA |
| |
|
| Source name |
ago3t2/ago3t3 (replicated 2)
|
| Organism |
Drosophila melanogaster |
| Characteristics |
Total RNA purified from whole ovaries from 2-4 day old flies.
|
| Growth protocol |
Adult 2-4 day old female flies were fed yeast paste and maintained at 25˚C.
|
| Extracted molecule |
total RNA |
| Extraction protocol |
The RNA was purified using Qiagen RNAeasy® kit following the manufacturer's protocol (Qiagen)
|
| Label |
Biotin
|
| Label protocol |
GeneChip® WT Double-Stranded DNA Terminal Labeling Kit (Affymetrix)
|
| |
|
| Hybridization protocol |
Recommended Affymetrix protocol
|
| Scan protocol |
The tiling arrays were scanned using an Affymetrix GeneChip® Scanner 3000 7G according to manufacturer's instructions.
|
| Description |
Processed data: The first column is the transposon ids from flybase or transcript ids from RefSeq. The second column is the the summarized/normalized expression value.
|
| Data processing |
Since transposons account for more than 10% of the probes on the tiling array, a special normalization workflow has been taken to avoid over-normalization of signals. Coordinates and raw signal values for Drosophila tiling 2.0R array probes were extracted from Affymetrix Tiling Array Software. Then probes mapped to transposon regions were identified. The remaining non-transposon probes were quantile-normalized across wild type and mutant replicates based on the assumption that the overall signal distribution of these probes should remain the same across the different strains. Then the signals for transposon probes were calculated by looking up the normalized values of non-transposon probes at same raw signal level. To summarize the signal for each transposon element, probes mapped to any copy of the element were grouped together and Hodges-Lehmann estimator (ref1) was used to calculate the pseudomedian of their normalized signals. We used pseudomedian as it is less sensitive to the large number of outlier probes (probes with value of 1) in the tiling array experiments. Differentially expressed transposon elements were identified by contrasting their pseudomeidan values in three mutant replicates against three wild type replicates. To correct for multiple testing, False Discovery Rates (FDRs) were calculated from t-test P-values. FDR<0.02 were used to call significantly changed transposons. Transposons with average expression value less than a minimal expression threshold of 149 (the 50th percentile expression value of RefSeq mRNAs) in both wild type and mutants were flagged as they are at the detection limit of the tiling array. Expression values of mRNAs were summarized by calculating the pseudomedian of probes mapped to each of the RefSeq mRNA transcripts. Differentially expressed transcripts were identified using FDR<0.02 cutoff.
|
| |
|
| Submission date |
Jan 12, 2009 |
| Last update date |
May 03, 2009 |
| Contact name |
Hualin Xi |
| E-mail(s) |
[email protected]
|
| Organization name |
Boston University
|
| Street address |
44 Cummington Street
|
| City |
Boston |
| State/province |
MA |
| ZIP/Postal code |
02215 |
| Country |
USA |
| |
|
| Platform ID |
GPL6629 |
| Series (1) |
| GSE14370 |
Characterization of expression changes in armi,rhino,aub,ago3 mutants by tiling array |
|
