minDeltaPsi default about drop HOT 4 CLOSED

This parameter removes junctions with no (or little) variability as, anyway, no outliers will be detected on them, and the fit will be faster. I updated the default of drop init to 0.05

from drop.

Is it possible if only 1 sample out of 100 samples is an outlier that mindeltaPsi - 0.05 will miss this abnormal sample splicing? That is an important situation not to miss for rare disease research.

from drop.

Yes, if all samples has a deltaPsi of 0 and 1 sample has a deltaPsi of 0.05, it can be called as an outlier. Nevertheless, a splicing defect of 0.05 might not be strong enough to cause a disease. Our deltaPsi cutoff for a junction to be significant is 0.3. Nevertheless, we left them as parameters so that users can adjust them as they prefer.

from drop.

I see. I was confused how mindeltaPsi is calculated. If it was an average across all samples, then one sample wouldn't be detected, but I guess it is calculated relative to single samples and not an average across all samples.

from drop.