search

HAN-Siyu / LION

An Integrated R Package for Effective Prediction of ncRNA- and lncRNA-protein Interaction
https://doi.org/10.1093/bib/bbac420
GNU General Public License v3.0
9 stars 3 forks source link

Some issues I can't yet fix: #4

Open ArthurGasparindo opened 1 year ago

ArthurGasparindo commented 1 year ago

Essentially I've been working with the lncPro function of the LION package for a couple of months, and I even managed to get results and visualize them in cytoescape. I had some problems with my R and had to remove it from my machine and install everything again, but now some issues I've never encountered before started appearing (with the same data as before):

Error in checkForRemoteErrors(val): 6 nodes produced errors; first error: argument of length 0

Segmentation fault (core dumped)

warning: pf_Scale too large (several times, as if for all sequences) #This I've researched a bit and it seems to be a commom issue with ViennaRNA, but I have no ideia how to alter pf_scale through R.

I have a very high amount of sequences that the computer may be struggling with given its RAM memmory, but I've segmented the data into 10 files, and the issues persisted, so I tried running only 100 sequences just to verify if it could be the memmory that is causing the issues, but they kept happening.

HAN-Siyu commented 1 year ago

Hi @ArthurGasparindo ,

Thanks for using LION. I have just updated LION to v0.3.0.0, and in this version you can pass additional arguments to RNAsubopt and Predator. When you run run_lncPro(), using args.RNAsubopt to adjust pfScale parameter may help. Please let me know if you notice further issues with the updated version, though I can only help with the part of LION side...

Best, Siyu

HAN-Siyu commented 1 year ago

By the way, have you tried to use fewer cores to run the code? maybe running with 6 nodes consumed too much memory? The program processes sequences one by one. And if parallel.cores=6, R processes 6 sequences at the same time until all loaded sequences are processed. So only segmenting data into smaller files may not reduce memory usage much.

ArthurGasparindo commented 1 year ago

Hello, thanks for the answer. This helps immensely. I was wondering, to use the argos.RNAsubopt, how do I do this inside of the run_lncpro() function? Because in the Manual this argos,RNAsubopt is only present inside run_RNAsubopt() function?

Em sáb., 30 de set. de 2023 às 20:01, Han, Siyu @.***> escreveu:

By the way, have you tried to use fewer cores to run the code? maybe running with 6 nodes consumed too much memory?

— Reply to this email directly, view it on GitHub https://github.com/HAN-Siyu/LION/issues/4#issuecomment-1741877586, or unsubscribe https://github.com/notifications/unsubscribe-auth/AYBMZ4YTTQZWUSD2DVC3VVLX5CQFFANCNFSM6AAAAAA5NYOHO4 . You are receiving this because you were mentioned.Message ID: @.***>

HAN-Siyu commented 1 year ago

Hi @ArthurGasparindo,

The support of args.RNAsubopt has been implemented in functions run_lncPro(), runRNAsubopt(), and featureStructure(). Here I can see args.RNAsubopt in run_lncPro(). Please check if you can see it in the latest documentation (pdf manual page 44, or ?LION::run_lncPro in R). Thanks.

ArthurGasparindo commented 1 year ago

Thanks, this helped a lot, and I managed to run my first couple of files with no more issues, but then, in the next file I found an error in predator "ERROR: Only one input file is allowed", I tried to add the args.Predator = "-iSeqId" and then other arguments from the predator manual, but the error persisted. I found it weird that this error didn't happen before, since all my data is divided into 31 chunks of the same number of sequences and there was no difference in input format from one to the other, yet this error persists in some data chunks and other not.