gender: F age: 74 hyb_slide: S25 labeling_plate: P2 molecular_subtype: MS1a tumor_stage: Ta tumor_grade: G1 progression_to_higher_stage_or_radiotherapy_(yes/no): yes (T2) dod_excluded_(yes/no): yes
Treatment protocol
Urothelial carcinoma samples were collected by cold-cup biopsy from the exophytic part of the bladder tumor of patients undergoing transurethral resection at hospitals within the southern healthcare region of Sweden.
Extracted molecule
total RNA
Extraction protocol
Total RNA was extracted using Trizol reagent followed by purification on RNeasy micro columns (Qiagen) according to the manufacturers recommendations
Label
biotin
Label protocol
Total RNA was labeled using the Illumina TotalPrep-96 RNA Amplification Kit (Ambion), following the manufacurers instructions.
Hybridization protocol
According to the manufacturers specifications
Scan protocol
According to the manufacturers specifications
Data processing
Control probe filtering and background correction was performed within the BioArray Software Environment (BASE). The data was corrected for bias introduced by two different extraction methods (RNeasy and AllPrep) by a probe-by-probe rescaling of each sample, determining the scaling factor using the median value of RNeasy samples vs the median value of AllPrep samples. The smaller sample group (AllPrep) was rescaled to the same median as the larger group (DNeasy). Using the same method, we rescaled all probes of labeling plates P1, P2, P5, and P6 to have the same median values as P4 samples, thereby correcting for an observed labeling plate bias. Following bias correction, the data was quantile normalized using the limma package (Smyth, Stat Appl Genet Mol Biol. 2004) and converted to the log2 scale. The probes were annotated to HGNC GeneSymbols using the Illumina platform annotation file and probes without gene symbol were discarded. Next, 50% of the probes with lowest intensity were discarded and the remaining probes were mean-centered. Probes annotated with the same gene symbol were merged using the median value for each gene symbol. Normalized data in Sample tables below: Original detection p-values, along with expression values arrived at after: background correction, probe-by-probe rescaling adjustment for extraction method, probe-by-probe rescaling adjustment for labeling plate, quantile normalization, Log2 conversion, application of a 50% intensity filter and mean centering of probes. Matrix non-normalized (linked as supplementary file on Series record): The raw intensity values subjected only to background correction.
