-
Notifications
You must be signed in to change notification settings - Fork 30
QualityControl
Eaton edited this page Jan 15, 2019
·
1 revision
Several quality control metrics are applied in AltAnalyze to identify putative outlier samples. The main output of all of these analyses are PDF and PNG image files output to the results directory under the folder DataPlots
. A detailed description of these methods is provided here.
- Distribution of normalized log2 probeset intensity values.
- Mean raw signal intensities of each array.
- Mean absolute deviation (MAD) of the RMA residuals for each array.
- Distribution of log2 read-counts (exon and junction).
- Feature-level box-plots for the distribution of exon, junction and intron read-counts.
- Total number of reads for each sample, broken down by exon, junction and intron aligning.
- Hierarchical clustering of significant genes and outliers.
- Principal Component Analysis of samples (significant genes).
- Cell/tissue characterization of each sample (LineageProfiler).