search

bartongroup / RATS

Relative Abundance of Transcripts: An R package for the detection of Differential Transcript isoform Usage.
MIT License
32 stars 1 forks source link

More sensitive default settings? #50

Closed fruce-ki closed 5 years ago

fruce-ki commented 6 years ago

Following discussions, it has been made obvious to me that an effect size threshold of 0.20 is pretty high, especially in genes with just 2 isoforms, and that many plausible and potentially interesting DTU events lie marginally below that threshold.

Although the current default settings give decent results overall, it is also evident that there is room for improvement. It may be worth setting a stricter p-value cutoff as default and, instead, extending the effect size down to 0.10 and lowering the reproducibility thresholds.

fruce-ki commented 5 years ago

I decided to stick with the conservative threshold for the effect size. People who want more sensitive and noisy results should do it explicitly.