assemble could not start

[galstriker]

when I want to test tormes, it stop with this :
2020-04-24 17:08:56 - MEGAHIT v1.2.9
2020-04-24 17:08:56 - Using megahit_core with POPCNT and BMI2 support
Traceback (most recent call last):
File "/root/miniconda3/envs/tormes-1.1/bin/megahit", line 1038, in
main()
File "/root/miniconda3/envs/tormes-1.1/bin/megahit", line 988, in main
build_library()
File "/root/miniconda3/envs/tormes-1.1/bin/megahit", line 264, in checked_or_call
func(*args, **kwargs)
File "/root/miniconda3/envs/tormes-1.1/bin/megahit", line 723, in build_library
create_fifo('pe1', i, inpipe_cmd(opt.pe1[i]))
File "/root/miniconda3/envs/tormes-1.1/bin/megahit", line 710, in create_fifo
os.mkfifo(fifo_path)
PermissionError: [Errno 1] Operation not permitted

but all dependences are installed , megahit can pass selftest, whats the problem? Thanks

[nmquijada]

Hi @galstriker
First of all, in order to check that TORMES is working properly despite MEGAHIT, can you run the same analysis by using SPAdes as assembler instead? (enable the --assembler spades option. Note that it is not compatible with the --fast option)
Thanks,
Narciso

[galstriker]

Hi@nmquijada
Thanks for answering my question. I ran the pipeline with spades and it could assemble and all the other program could run. However, it stopped at making reports and gave me the following message:

Quitting from lines 203-212 (tormes_report.Rmd) Error in file(file, "rt") : cannot open the connection Calls: <Anonymous> ... withCallingHandlers -> withVisible -> eval -> eval -> read.table -> file In addition: Warning messages: 1: package 'ggplot2' was built under R version 3.6.2 2: package 'knitr' was built under R version 3.6.2 3: package 'plotly' was built under R version 3.6.3 4: package 'reshape2' was built under R version 3.6.3
The R packages were installed when I ran tormes-setup, so I could not figure out what the problems is.

[biobrad]

Hi galstriker, When did you install Tormes? Narciso has been releasing a few fixes recently and I think one from not too long ago addressed issues found with 'R' versions.

[galstriker]

Hi galstriker, When did you install Tormes? Narciso has been releasing a few fixes recently and I think one from not too long ago addressed issues found with 'R' versions.

Hi biobrad, I installed tormes a few days ago, following the instruction.

[nmquijada]

Hi @galstriker
Good that TORMES worked by using SPAdes.
Let's see your report problem. Can you please share the "tormes_report.Rmd" file? It would be inside the "report_files" directory, so first do:
tar xzf report_files
Ignore the warnings with the R packages, they won't affect their activity.

Regarding your problem with MEGAHIT, I still don't figure out what can it be, although I am afraid it migh not be related with the environment installed.
Can you please try the following command? (modify it accordingly) Acivate tormes-1.1 environment first:
megahit -1 <R1 READ> -2 <R2 READ> -o <OUTPUT DIRECTORY> -t <#CPUS> --no-hw-accel

Best,
Narciso

[galstriker]

Hi @galstriker
Good that TORMES worked by using SPAdes.
Let's see your report problem. Can you please share the "tormes_report.Rmd" file? It would be inside the "report_files" directory, so first do:
tar xzf report_files
Ignore the warnings with the R packages, they won't affect their activity.

Regarding your problem with MEGAHIT, I still don't figure out what can it be, although I am afraid it migh not be related with the environment installed.
Can you please try the following command? (modify it accordingly) Acivate tormes-1.1 environment first:
megahit -1 <R1 READ> -2 <R2 READ> -o <OUTPUT DIRECTORY> -t <#CPUS> --no-hw-accel

Best,
Narciso

Hi @nmquijada ，
tormes_report.zip
here is the .Rmd file. I ran the render_report.sh it still reported error and stop at about 45%.

[nmquijada]

Hi @galstriker
If there's an error in the .Rmd file or in the file that it is looking for, it will return the same error. This .Rmd can be manually edited and re-run by using the render_report.sh, in order to generate a new report. But it doesn't fix bugs. If you see the error you sent me, it says "Quitting from lines 203-212 (tormes_report.Rmd)". That lines in the tormes_report.Rmd correspond to the 2nd MLST scheme for E. coli.
Can you share the entire "report_files" directory so I can inspect the files? If you think you have sensitive data, we can find another way for sharing.
Thanks,
Narciso