generateEvents missies some Retained Intron (RI) events

[HegedusB]

Hi.
I have tried the the SUPPA's generateEvents function and I have found that in some case intron retention was not opserved or are missing from the output ioe files. Here is an example:

used command:
python3 /SSD2/bhegedus/installs_dir/Suppa/suppa.py generateEvents -i goi.gtf -o test.events -e SE SS MX RI FL -f ioe

used annotation:

Can you help me to understand the problem.
Cheers,
Botond

[EduEyras]

Hi, this expected by the definition of the RI event In the default mode, the RI is quite restrictive and it expects that the exon retaining the intron also share the external splice sites of the exons flanking the introns You can relax that condition with the "variable boundary" option: -b V You can parametrise it to account for variable distances to the flanking exons with -t Try generating the RI events with e.g -b V -t 50 (or a higher value) I'd expect that would pick up the case you're showing I hope this helps E.

…

On Thu, 24 Nov 2022 at 20:03, HegedusB ***@***.***> wrote: Hi. I have tried the the SUPPA's generateEvents function and I have found that in some case intron retention was not opserved or are missing from the output ioe files. Here is an example: [image: image] <https://user-images.githubusercontent.com/9871809/203739042-b80a514e-7257-48fb-9c4f-acdcceb43915.png> used command: python3 /SSD2/bhegedus/installs_dir/Suppa/suppa.py generateEvents -i goi.gtf -o test.events -e SE SS MX RI FL -f ioe used annotation: [image: image] <https://user-images.githubusercontent.com/9871809/203739905-008343eb-1d77-48b7-9151-c6d8ad746263.png> Can you help me to understand the problem. Cheers, Botond — Reply to this email directly, view it on GitHub <#156>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADCZKB5ZNYTNOLKLRGUPRXTWJ4VOXANCNFSM6AAAAAASKBJXZA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[HegedusB]

Hi,
Thank you for the clear explanation. I have tried with the mentioned parameter changes. Now it can locate the RI event correctly but identify an extra artificial A3 event.

So, if I want to avoid the appearance of artificial events, I should use the strict approach.
Cheers,
Botond

[EduEyras]

You can run it only for RI events, which may be probably the case where you want to relax the constraints For all other event types, you can keep the restrictive definition E.

…

On Fri, 25 Nov 2022 at 20:53, HegedusB ***@***.***> wrote: Hi, Thank you for the clear explanation. I have tried with the mentioned parameter changes. Now it can locate the RI event correctly but identify an extra artificial A3 event. [image: image] <https://user-images.githubusercontent.com/9871809/203952854-930e34fc-d035-484f-86ed-febd1674453e.png> So, if I want to avoid the appearance of artificial events, I should use the strict approach. Cheers, Botond — Reply to this email directly, view it on GitHub <#156 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADCZKBY5TKP6RQX4F7TOKFDWKCEDNANCNFSM6AAAAAASKBJXZA> . You are receiving this because you commented.Message ID: ***@***.***>

[HegedusB]

Thanks! This is a great idea. I will do like that.
Cheers,
Botond

[Dongmeng-wang]

Hi Botond, thank you so much for bringing up this question. Can I ask what value you finally set for -t, 50 or a higher value? Thanks!