Add RNA 2D structure to RNA pages

[kimrutherford]

This should be possible with the r2dt-web web component :

• https://github.com/RNAcentral/r2dt
• https://github.com/RNAcentral/r2dt-web
• https://r2dt.readthedocs.io/en/latest/widget.html

[ValWood]

SGD have just done this 👍 https://www.yeastgenome.org/blog/sgd-newsletter-summer-2023 Maybe we should add sooner.

[ValWood]

This isn't super urgent but let's put it at high priority because it looks super nice, and is probably quick (ish) to do.

[kimrutherford]

Try to embed as a widget on the tRNA, snRNA, snoRNA and rRNA pages (only). It should be the default widget for those RNAs.

[kimrutherford]

SGD have just done this 👍 https://www.yeastgenome.org/blog/sgd-newsletter-summer-2023

As far as I can see, the SGD pages have only schematic 2d RNA structure:

https://www.yeastgenome.org/locus/YNCQ0010W#overview

Is that what we want?

The images are just embedded links to RNAcentral so would be easy to add to our gene pages:

https://rnacentral.org/api/v1/rna/URS00004C1880/2d/svg/

[ValWood]

they have the interactive structure on the "sequence" pages: https://www.yeastgenome.org/locus/YNCQ0010W/sequence#secondary_structure

[kimrutherford]

they have the interactive structure on the "sequence" pages:

Ah, thanks!

[kimrutherford]

I've had a look at the r2dt-web widget. I can't see any configuration options for making it smaller. I'm about to open some issues on their GitHub repo.

We also need to remove some of the text above and below.

All of this:

and this:

is part of the widget

That's why the SGD page has the text "Secondary Structure" twice:

[ValWood]

Why do we need to remove the key and the other bits (don't want to repeat the title but presumably the other parts are useful?)

[kimrutherford]

Why do we need to remove the key and the other bits (don't want to repeat the title but presumably the other parts are useful?)

If we match the other widgets then we only have a limited amount of vertical space - about 400 to 500 pixels . The key on its own is 200 pixels high.

[ValWood]

right yes, that's an issue for the widget...

[kimrutherford]

I decided to just add the full 2D structure widget to the RNA pages because it was easier than making a proper widget. It takes up a bit of space but it will probably do for now:

https://www.pombase.org/gene/SPRRNA.54

[kimrutherford]

I noticed that this structure isn't available:

https://www.pombase.org/gene/SPRRNA.01 https://rnacentral.org/rna/URS0001BBE0D5/4896

[ValWood]

I think its greayt like that. The RNA pages don't have much else on them anyway

[ValWood]

I noticed that this structure isn't available:

is this something to do with it being mitochondrial genome?

[kimrutherford]

is this something to do with it being mitochondrial genome?

I tried running r2dt manually (https://rnacentral.github.io/r2dt-web/) with the sequence of SPRRNA.01 but it returned "The sequence did not match any of the templates."

I don't know what that means. :-)

[ValWood]

So here are we trying to get structures for all sn/sno,tRNA/rRNA? If so, it is possible that SPRRNA.01 is excluded because it possibly doesn't have a recognisable 2-d structure...

[ValWood]

There is no 2-structure on the RNA central page either

[kimrutherford]

So here are we trying to get structures for all sn/sno,tRNA/rRNA?

We're showing the widget for the types in your list ("tRNA, snRNA, snoRNA and rRNA pages (only)") but only if the RNA has an ID like URS0001BBE0D5 in Chado (because that's the ID that the widget uses).

[kimrutherford]

If so, it is possible that SPRRNA.01 is excluded because it possibly doesn't have a recognisable 2-d structure...

I tried the sequence in a different RNA structure predictor. The result is quite funky. :-)

http://rna.tbi.univie.ac.at//cgi-bin/RNAWebSuite/RNAfold.cgi?PAGE=3&ID=rliF4Fn77w

[ValWood]

It is, I suspect it is probably processed before it is folded. it's very long 2827 nt

[kimrutherford]

Decision from Zoom: change to widget

[kimrutherford]

There are a bunch of snoRNAs that have no structure. Here are a sample: https://www.pombase.org/gene/SPSNORNA.18 https://www.pombase.org/gene/SPSNORNA.04 https://www.pombase.org/gene/SPSNORNA.02 https://www.pombase.org/gene/SPNCRNA.498

For others the structures are a bit unconvincing: https://www.pombase.org/gene/SPNCRNA.7474 https://www.pombase.org/gene/SPSNORNA.23

[kimrutherford]

Decision from Zoom: change to widget

I've done that and deployed the change to pombase.org. It's usable as is but I'm still tweaking things.

[ValWood]

In this example, when you click on the structure, the rest of the page disappears: https://www.pombase.org/gene/SPSNORNA.18

[kimrutherford]

I get this message because it has no structure. Is that what you mean?

[ValWood]

There are a bunch of snoRNAs that have no structure.
Here are a sample:
https://www.pombase.org/gene/SPSNORNA.18
 This one looked odd becasue it was shsort. I have extended it to 74 nt. It will probable get a stucture in the futire

https://www.pombase.org/gene/SPSNORNA.04
https://www.pombase.org/gene/SPSNORNA.02
https://www.pombase.org/gene/SPNCRNA.498

Have no structrue in RNACentral ~(on the whole I would not expect snoRNAs to have folded structures because they mainly have base pairing with other RNAs rather than internal base pairing~

For others the structures are a bit unconvincing:
https://www.pombase.org/gene/SPNCRNA.7474
This is the primary "unprocessed transcript" We need to think of a better way to annotate this....

https://www.pombase.org/gene/SPSNORNA.23 Yes this looks odd...

[ValWood]

I get this message because it has no structure. Is that what you mean?

I see now that the rest of the page its there, but there is a big blank section

[ValWood]

[kimrutherford]

Yep. Because of the way it's implemented, the space used by the widget is fixed.

Perhaps we should query RNACentral to find which RNAs have structures? That way we can insert the widget only if there's a structure to show.

on the whole I would not expect snoRNAs to have folded strucures

Should we not show the widget for snoRNAs in that case?

[ValWood]

Ignore that, the snoRNA's are probably OK

[ValWood]

Yeah maybe query for structures...a pain...

[kimrutherford]

Yeah maybe query for structures...a pain...

No problem. It's only a little bit of a pain. :-)

Once we have a script that does the querying, re-running to update the data will be simple.

[kimrutherford]

Yeah maybe query for structures...a pain...

Turns out it's easy to query RNACentral for pombe RNAs that have a structure. Well done them!

From here: https://rnacentral.org/expert-database/pombase Click on "Browse all PomBase sequences" to get to this list: https://rnacentral.org/search?q=expert_db:%22PomBase%22

Then tick the "Secondary structure: available" checkbox to get: https://rnacentral.org/search?q=expert_db:%22PomBase%22%20AND%20has_secondary_structure:%22True%22

Excellent stuff!

Here's the list: https://www.pombase.org/results/from/id/68cd3116-9daa-40be-8ece-0a62cc6f2964

I've add a file to SVN with this information:

pombe-embl/supporting_files/rna_genes_with_2d_structure.tsv

[kimrutherford]

Perhaps we should query RNACentral to find which RNAs have structures? That way we can insert the widget only if there's a structure to show.

From Friday morning we'll only show the widget if the structure is available at RNAcentral. That's a better user experience.

[ValWood]

I will write a short announcement /news item for tomorrow.

Didi you try to hide the error?

[kimrutherford]

Yep, that's what I've fixed: from tomorrow the widget will be displayed only on RNA pages where there is a structure available from RNAcentral. So no more errors from tomorrow.

[ValWood]

I meant hide the title!

[ValWood]

draft

We have added RNA secondry structure predictions using the RTDT service provided by RNACentral The predictions appear as an option in the summary section of the page e.g. https://www.pombase.org/gene/SPRRNA.03 You can retrieve the list of all RNA genes with secondary structure predictions using the a “canned” query under “commonly used queries” here https://www.pombase.org/query

Do you want to edit and send to pombelist and make a suitable version for the news item?

[kimrutherford]

I've added a reminder to the PomBase calendar to update the file of genes with RNA structure predictions. It's currently every two months but I can do it more regularly if think that's needed.

[ValWood]

they are unlikely to change. 6 monthly (or even annually) would probably be fine...

[kimrutherford]

they are unlikely to change. 6 monthly (or even annually) would probably be fine...

It's now three times a year. I sometimes miss these things in the calendar so I thought 3 times would be better than 1 or 2.

[kimrutherford]

I've added a new item. Let me know if you'd like any changes.

Is there anything else to do on this issue?

[ValWood]

no, all finished thanks!