Sign inSign up

trust1/jmml

By trust1

Updated almost 4 years ago

JMML classifier

Image
0

248

trust1/jmml repository overview

JMML classifier box

Container with R v3.1.3 on Ubuntu v22 based on Rocker base

Includes multiple libraries and packages for running JMML classifier build around RnBeads R library.

Step 1. Launch container (assuming docker is already installed on your system):

docker run \
-v /YOUR/LOCAL/PATH2/idats_in:/in \
 -v /YOUR/LOCAL/PATH2/out:/out \
 -v /YOUR/LOCAL/PATH2/reports:/reports \
 --name JMML -it trust1/jmml 

Where you have to modify directions to three places:

1 . /YOUR/LOCAL/PATH2/idats_in: -- local directory with methylation idat files. The source folder.

  1. /YOUR/LOCAL/PATH2/out: -- path on your system where to deposit beta values and textual reports.

  2. /YOUR/LOCAL/PATH2/reports: -- Local place where PDF and HTML reports will be generated.


Step 2. Launch R script. The idat processing and classifier script for individual sample can be initiated with this command:

Rscript /JMML.code/process_sample.R TESTMENAME 202292320091_R01C01

Where you have to change sample information accordingly:

TESTMENAME -- this is sample identifier (patient ID or whatever non blank string)

202292320091_R01C01 -- idat barcode name. Idat file need to be unzipped and stored in your local folder which you defined in [ /YOUR/LOCAL/PATH2/idats_in ] in step 1.


Single run in non interactive mode.

Container instance will be removed after job finish or crash.

docker run --rm \
-v /YOUR/LOCAL/PATH2/idats_in:/in \
 -v /YOUR/LOCAL/PATH2/out:/out \
 -v /YOUR/LOCAL/PATH2/reports:/reports \
  trust1/jmml \
 Rscript /JMML.code/process_sample.R \
TESTME1 \
202292320091_R01C01

Container on biowulf with singularity

Initiate interactive mode on Biowulf. Then priming with singularity:

/data/MDATA/NormRcode/JMML_classifier	
module load singularity
singularity pull docker://trust1/jmml:latest  

This need to be done only once. It makes container image in your location.

For routine run launch those lines in interactive mode:

module load singularity
cd /data/MDATA
export TMPDIR=/lscratch/$SLURM_JOB_ID
export SINGULARITYENV_THREADS=2
source /usr/local/current/singularity/app_conf/sing_binds

singularity run --bind /data/MDATA/compass/iScan_raw/207019360020/:/in,\
/data/turakulovr2/JMML_Mehtylation_Classifier-main/out/:/out,\
/data/turakulovr2/JMML_Mehtylation_Classifier-main/reports/:/reports \
/data/MDATA/NormRcode/JMML_classifier/jmml_latest.sif \
Rscript /JMML.code/process_sample.R DL20 207019360020_R07C01

Results will be generated in here:

/data/turakulovr2/JMML_Mehtylation_Classifier-main/reports


Credits:
  • Maximilian Schoenung
  • Julia Meyer
  • Rust Turakulov

Tag summary

Content type

Image

Digest

sha256:c24637e34

Size

1.5 GB

Last updated

almost 4 years ago

docker pull trust1/jmml