1
0
Entering edit mode
@82b47109
Last seen 9 days ago
Australia

Hi Alex, This is not an issue but rather a question about one of the parameters of SpliceWiz.

To get significant differential events, I saw in the quick start tutorial of SpliceWiz, we can obtain a list of differential events with delta PSI > 5%, res_limma.filtered <- subset(res_limma, abs(deltaPSI) > 0.05). My question is that if I use deltaPSI > 0.2 would that mean the differential events between condition B versus A are more confident than deltaPSI >0.05?

Code should be placed in three backticks as shown below

res_edgeR.filtered <- subset(res_edgeR, abs(deltaPSI) > 0.2 &
PValue< 0.05)

sessionInfo( )

SpliceWiz • 84 views
0
Entering edit mode
a.wong ▴ 10
@user-24863
Last seen 3 days ago
Australia

Hi Reza,

"Confidence" is often synonymous with statistical significance which is measured as p-values. In context of multiple testing, the adjusted p value is favored.

Sometimes, adding a filter of |delta PSI| is helpful to avoid including events with PSIs at the limit margins. E.g. if in control, PSI = {0.0001, 0.00009, 0.00011} and in treatment PSI = {0.0010, 0.0011, 0.0012}, it will be called as highly statistically significant although at the protein translation level this may be functionally meaningless. There is no set rule as to what |delta PSI| filter to use, but in my experience, > 0.2 is too high.

However, for cherry-picking "top" results then |delta PSI| > 0.2 might be useful.