You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
uorf html track
Some of the clearly active ORFs in this transcript are not predicted active.
All predicted uORFs by ORFquant are marked with uorf numbers, like uORF1 etc.
Check out uORF that starts at 601, what do you think is going on there?
Zoom in to the uORF at 600 in the HTML by holding left mouse and drag over region wanted.
I have checked through the settings, and do not see a clear reason why it is ignored, using default settings.
It does not share stop codon with already predicted ORF.
1% coverage remaining is default for ORF detection to stop, this ORF contains > 1% remaining reads.
It has far more than 10 p-site counts, which is minimum.
more than 50% of the reads are in frame
Can you see a reason?
The text was updated successfully, but these errors were encountered:
It might be that this uORF is very small, thus not making the cut for 3nt periodicity analysis.
If you can provide me with the data and details of your ORFquant run I can give you more feedback.
uorf html track
Some of the clearly active ORFs in this transcript are not predicted active.
All predicted uORFs by ORFquant are marked with uorf numbers, like uORF1 etc.
Check out uORF that starts at 601, what do you think is going on there?
Zoom in to the uORF at 600 in the HTML by holding left mouse and drag over region wanted.
I have checked through the settings, and do not see a clear reason why it is ignored, using default settings.
Can you see a reason?
The text was updated successfully, but these errors were encountered: