Fold Change

[fruce-ki]

RATs uses the difference between proportions as its effect size and the volcano plot uses that difference as its x axis. Fold-change was dismissed early on, as fold-cahnge of proportions is difficult to interpret and possibly nonsensical. However, fold-change of the transcript abundance could be useful additional information and possibly useful as an additional constraint.

• [x] Add FC field to the Transcripts table and calculate fold-change of abundance. Even if purely for descriptive/informative purposes, along with the other stats fields.
• [x] Plot the FC of DTU and non-DTU genes and see if there is scope for adding an fc constraint.
• [x] Provide conventional volcano: pval VS fc .
• [ ] Normalize FC?

[fruce-ki]

Fold-changes require normalisation of the input to cancel out sequencing depth variation. Even so, the FC will be influenced by DGE. I think having the FC field and the plots may still be useful in identifying outlier/interesting transcripts. However, without normalisation, a hard threshold included in the DTU decision process is not possible.