RNA type identification after graphclust analysis

[backofenlab]

RNA type identification after graphclust analysis

[mavino]

I am repeating an NSPDK_sparseVect process on galaxy locally installed, before I start the process I notice that the system does not recognize directly as input file in the field "gspan file" the file that I generated from the former process "fasta_to_gspan". This happened even when I ran the analysis on the other galaxy, in that case I managed to let the system recognize this file (do not remember how), even though the system reported it as hidden file (then, as I mentioned in my previous message, the analysis runs forever).The problem might rely on the generation of this file gspan file then...

[mavino]

Just realized that I miss the RNAfold step...running it and see what happens...

[mavino]

RNA fold gives me this error:

Fatal error: Exit code 1 ()
Warning from traverse_loop. Loop 2 has crossed regions
Warning from traverse_loop. Loop 2 has crossed regions
Warning from traverse_loop. Loop 5 has crossed regions
Warning from traverse_loop. Loop 5 has crossed regions
Warning from traverse_loop. Loop 3 has crossed regions
Warning from traverse_loop. Loop 2 has crossed regions
Warning from traverse_loop. Loop 2 has crossed regions
Unexpected large magnitude discriminant = -1.04881 0.494008

[mavino]

I broke down my initial fasta input file (which came from the whole GTF genome) in single chromosomes...the RNA fold analysis completed successfully this time...running the next steps of GraphClust_main_3r pipeline

[mavino]

I am closing this issue because I guess the problem was related with my initial big input file. When I run on shorter files everything goes smoothly.

[mmiladi]

Hi @mavino ,
I would suggest to try the motif_finder workflow as well. Depending on the question you ask, it might be a better option. You may also want check the graphclust server frontpage: https://graphclust.usegalaxy.eu/

[mavino]

Hi @mmiladi ,
thank you so much for your suggestion...
what I really want is to classify those kinds of RNA I get at the end.
It would be great to have a plot like the lower right panel of Fig. 1 of your graphclust2 paper, the one that tells you your candidate motifs and classify them in different types of RNA. However, I am seeing that at the end of my completed GraphClust_main_3r pipeline I do not get that plot or a table to tell me that...is the pipeline motif_finder going to give me that plot or something telling me that this particular cluster resembles a specific kind of RNA? Thank you so much!!

[mavino]

Any further thought on my last question?

[mavino]

Would it be a good idea to input the cluster sequnces found by graphclust into Rfam (https://rfam.xfam.org/search?q=mana#tabview=tab1) to detect the type of RNA?

[mmiladi]

Hi @mavino ,

From the error report you have posted, I suspect there should be an error with the input fasta file. I wrote you an email to look into the specific case.
Best,
Milad