ORF segmentation (AUG vs non-canonical starts)

[CDieterich]

We noticed in a head-to-head comparison of ORFs
predicted with canonical and non-canonical starts
that some longer ORFs with non-canonical starts
do not have any supporting reads in the extended 5'part.

We should think about on how to deal with this.

[m-swirski]

Yes, I have seen the same thing. In fact It happens for most ORFs once alternative codons are being searched for. Are there plans to fix this?

[eboileau]

This is on our to-do list and should be addressed soon. If you have any relevant examples, please don't hesitate to post these here.

[m-swirski]

I have looked closer into the profiles and it seems that the problem doesn't lie in 'no reads at all' but small fraction of them compared to other potential TIS (or annotated one).