search

GoekeLab / bambu

Reference-guided transcript discovery and quantification for long read RNA-Seq data
GNU General Public License v3.0
182 stars 22 forks source link

Errors and warnings explanation #416

Open zhangkn3 opened 6 months ago

zhangkn3 commented 6 months ago

Hi Andre, Hope everything goes well with you! I have been using Bambu for months. It is a really good one. However, I met these errors and warnings these days, and the job seems to be halted. But even with these errors and warnings, I can still get the results and novel transcripts from Bambu. Do I need to do something to solve the issues, or I can just ignore them? Here are my codes when running Bambu.

library(bambu) gencode_seq <- "/gpfs/fs1/data/hg38.fa" gencode_gtf <- prepareAnnotations("/gpfs/fs1/data/hg38.knownGene.gtf") uncorrected_bam_file <- "/gpfs/fs1/data/filtered_minimaped_GBM013.sort.bam" corrected_bam_file <- "/gpfs/fs1/data/filtered_corrected_minimaped_GBM013.sort.bam" ont_iso_analysis <- bambu(reads = c(uncorrected_bam_file, corrected_bam_file), annotations = gencode_gtf, genome = gencode_seq, ncore = 8) writeBambuOutput(ont_iso_analysis, path = "/gpfs/fs1/data/GBM013/") ont_iso_analysis.novel <- ont_iso_analysis[mcols(ont_iso_analysis)$novelTranscript,] writeBambuOutput(ont_iso_analysis.novel, path = "/gpfs/fs1/data/GBM013/novel/")

And, these are the issue reports. image

Looking forward to your response! Best, Kenan

andredsim commented 6 months ago

Hi Kenan,

If you are meaning the xgboost warnings, that is fine to ignore. We just need to update the packages and push the changes. We just keep getting distracted with other tasks but I will hopefully find the time to do this soon. Given that bambu() finishes running, and the remaining warnings are not result impacting I think its fine to ignore for now. I believe the finally warnings must be happening during the writeBambuOutput() function call. But as you are still getting your outputs I think that should be fine too.

Do you still get the error: subscript contains invalid names If you do not run this line ont_iso_analysis.novel <- ont_iso_analysis[mcols(ont_iso_analysis)$novelTranscript,] and instead output the whole analysis? writeBambuOutput(ont_iso_analysis, path = "/gpfs/fs1/data/GBM013/novel/") I am wondering if perhaps the filtering is doing something we didn't forsee.

Best of luck with your future runs of Bambu

Kind Regards, Andre Sim

zhangkn3 commented 6 months ago

Thanks Andre, I will run the code and come to you later.