Question

using DiscoSNP++ gives error concerning $TERM and -T

0

Entering edit mode

2.7 years ago

paulien.philippe • 0

Hi everyone,

I'm trying to run DiscoSNP++ on data from the Tara Oceans Expedition but I ran into a problem that i have trouble solving. The input I'm using are two very large .gz files (study accession PRJEB4352). The output that I need are a .fa file and a .vcf file, but those don't show up. This is the script I used for the job which took about 40 minutes to complete:

#!/usr/bin/env bash

#PBS -l nodes=1:ppn=1
#PBS -l walltime=02:00:00
#PBS -l pmem=8gb
#PBS -A default_project

cd /vsc-hard-mounts/leuven-data/341/vsc34135/Laso-Jadart
source /vsc-hard-mounts/leuven-data/341/vsc34135/miniconda3/etc/profile.d/conda.sh
conda activate discosnp_env

curl -o ERR868369_1.fastq.gz ftp://ftp.sra.ebi.ac.uk/vol1/fastq/ERR868/ERR868369/ERR868369_1.fastq.gz
curl -o ERR868369_2.fastq.gz ftp://ftp.sra.ebi.ac.uk/vol1/fastq/ERR868/ERR868369/ERR868369_2.fastq.gz
run_discoSnp++.sh -r fof_1.txt –k 51 -b 1 -T

With fof_1.txt being: ERR868369_1.fastq.gz ERR868369_2.fastq.gz

This is part of the error job output

And this is part of the job output:

I'm quite new to making scripts and working with the VSC so any help is appreciated.

DiscoSNP VSC bash • 1.2k views

ADD COMMENT • link updated 2.4 years ago by Ram 44k • written 2.7 years ago by paulien.philippe • 0

score 0 · Answer 1 · 2022-03-23

0

Entering edit mode

2.7 years ago

pierre.peterlongo ▴ 900

Hi there.

The tput error is due to the fact that the TERM variable is not set in your case (which is the terminal to be used). (See stack_overflow_post)

It may be solved setting this variable before to run the discoSnp script.

TERM=xterm
run_discoSnp++.sh -r fof_1.txt –k 51 -b 1 -T

Please let me know if this also solves the graph construction issue.

Also be sure that fof_1.txt contains two lines:

ERR868369_1.fastq.gz
ERR868369_2.fastq.gz

Depending on how you which to consider these two files please read the doc (section Input File of file format)

I hope this helps,

Pierre

ADD COMMENT • link 2.7 years ago by pierre.peterlongo ▴ 900

0

Entering edit mode

Hi Pierre.

Thanks for the fast response. I've tried out your solution together with moving to a smaller dataset and adding some extra "just to be safe" code. It doesn't seem to work though. I still get the same errors... Here's my tweaked script

#!/usr/bin/env bash

#PBS -l nodes=1:ppn=1
#PBS -l walltime=02:00:00
#PBS -l pmem=4gb
#PBS -A default_project

cd /vsc-hard-mounts/leuven-data/341/vsc34135/Ofunato
source /vsc-hard-mounts/leuven-data/341/vsc34135/miniconda3/etc/profile.d/conda.sh
conda activate discosnp_env

curl -o ofunato1.fastq.bz.1 https://ddbj.nig.ac.jp/public/ddbj_database/dra/fastq/DRA005/DRA005744/DRX084576/DRR090871_1.fastq.bz2
bzcat ofunato1.fastq.bz.1 | gzip -c >ofunato1.1.gz
curl -o ofunato1.fastq.bz.2 https://ddbj.nig.ac.jp/public/ddbj_database/dra/fastq/DRA005/DRA005744/DRX084576/DRR090871_2.fastq.bz2
bzcat ofunato1.fastq.bz.2 | gzip -c >ofunato1.2.gz

TERM=xterm
run_discoSnp++.sh -r ofunato_1.txt -T

with ofunato_1.txt:

ofunato1.1.gz
ofunato1.2.gz

The job takes less than a minute to run, so it must be something in the very beginning that goes wrong.

ADD REPLY • link 2.7 years ago by paulien.philippe • 0

0

Entering edit mode

Hello Paulien

I've no issue with the master and the last released (2.6.2) versions (on an osx machine), using exactly your command lines.

###############################################################
 #################### DISCOSNP++ FINISHED ######################
 ###############################################################
DiscoSnp++ total time in seconds: 277
################################################################################################################
 fasta of predicted variant is "discoRes_k_31_c_3_D_100_P_3_b_0_coherent.fa"
 Ghost VCF file (1-based) is "discoRes_k_31_c_3_D_100_P_3_b_0_coherent.vcf"

Can you make a try with the last release https://github.com/GATB/DiscoSnp/releases (you may either install the sources or use any of the precompiled versions mac or Linux)

Depending on your feedback we will update the conda package.

I hope this helps.

ADD REPLY • link 2.7 years ago by pierre.peterlongo ▴ 900

0

Entering edit mode

Hi Pierre,

I tried installing the latest version via the sources but I get several error messages.

git clone --recursive https://github.com/GATB/DiscoSnp.git

gives me:

fatal: reference is not a tree: 7a2202e751a11a5a7125eb756b6ab2d285fe5f5f
Unable to checkout '7a2202e751a11a5a7125eb756b6ab2d285fe5f5f' in submodule path 'thirdparty/gatb-core'

and then it just stops and goes back to the command line. I continued with

sh INSTALL

Which gives me several errors:

fatal: reference is not a tree: 7a2202e751a11a5a7125eb756b6ab2d285fe5f5f
Unable to checkout '7a2202e751a11a5a7125eb756b6ab2d285fe5f5f' in submodule path 'thirdparty/gatb-core'
INSTALL: line 18: cmake: command not found
make: *** No targets specified and no makefile found.  Stop.
Running simple test...


 Running discoSnp++ 2.3.X, in directory /vsc-hard-mounts/leuven-data/341/vsc34135/Ofunato/DiscoSnp with following parameters:
         read_sets=fof.txt
         prefix=discoRes_k_31_c_3
         c=3
         C=2147483647
         k=31
         b=0
         d=1
         D=100
         s=
         P=3
         p=discoRes
         G=
         e=
         x=
         starting date=Wed Apr 13 12:03:54 CEST 2022


../run_discoSnp++.sh: line 503: /vsc-hard-mounts/leuven-data/341/vsc34135/Ofunato/DiscoSnp/build/bin/read_file_names: No such file or directory
 ############################################################
 #################### GRAPH CREATION  #######################
 ############################################################
/vsc-hard-mounts/leuven-data/341/vsc34135/Ofunato/DiscoSnp/build/ext/gatb-core/bin/dbgh5 -in fof.txt_discoRes_k_31_c_3_D_100_P_3_b_0_removemeplease -out discoRes_k_31_c_3 -kmer-size 31 -abundance-min 3 -abundance-max 2147483647 -solidity-kind one -verbose 1 -skip-bcalm -skip-bglue -no-mphf -histo-max 1000000
../run_discoSnp++.sh: line 521: /vsc-hard-mounts/leuven-data/341/vsc34135/Ofunato/DiscoSnp/build/ext/gatb-core/bin/dbgh5: No such file or directory
 there was a problem with graph construction$ reset
diff: discoRes_k_31_c_3_D_100_P_3_b_0_coherent.fa: No such file or directory
*** Test: FAILURE on diff .fa

I honestly don't really know how to install DiscoSNP using a precompiled version.

PS: is it right that I don't get the latest version using conda? I get version 1.5.3.

ADD REPLY • link 2.6 years ago by paulien.philippe • 0

score 0 · Answer 2 · 2022-04-19

fatal: reference is not a tree: 7a2202e751a11a5a7125eb756b6ab2d285fe5f5f

This is due to your git version.

With git 1.8.3.1 I have the error
With git 2.19.1 this is fine.

About the rest of your errors:

Things could not have compiled as all modules were not downloaded
INSTALL: line 18: cmake: command not found:
- You need cmake to compile this discoSnp (as most of the tools).

score 0 · Answer 3 · 2022-04-29

Hello Pierre,

I managed to run DiscoSnp++ now! It had to do with my conda environment that wasn't properly set. My apologies for the hassle anyway.

I noticed that you cowrote metaVaR, so I figured I could ask you my question(s) about that matter too. So the preprocessing step doesn't seem to work. I uploaded the perl script of metavarFilter.pl and ran it on my new .vcf file with the -a option set to 5. It returns empty .txt files so something get wrong. The only message I get is

Use of uninitialized value $d[3] in pattern match (m//) at metavarFilter.pl line 47, <IN> line x.

x going from 1-20. So it seems like the d[3] value needs to be defined/initialized but I don't know what it refers to.

Thanks in advance for all the help! Paulien