Produce a ggplot object containing a plot of residual standard deviation against mean count.
plot_stability(y, x = NULL, design = NULL, bins = 20)
| y | Transformed counts matrix. |
|---|---|
| x | Optional, original counts matrix. |
| design | Matrix specifying a linear model with which to calculate residuals. |
| bins | Number points in the graph. |
A ggplot object.
This must be print()-ed to actually plot.
Genes are partitioned evenly into "bins" bins by average expression level. Mean residual standard deviation is plotted against mean count.
If the variance stabilizing transformation has been successful, this plot should be close to a horizontal line. However it is normal for the standard deviation to drop off for counts below 5.