bchobrut/vdj repository overview

⁠Use

Run the docker image. -v specifies the volume to mount on the host. Replace the "/path/to/bams" part with that path. Leave "/mnt" as is. By default the VDJ recovery scripts will look for bams in the root of the mounted folder and will make a new folder there called "results" for the output

final results will be output to "VDJ_Recoveries.xlsx" in the results folder. A copy in hdf format will also be written as "vdj_recoveries.h5"

On singularity:

⁠Advanced Options

For advanced features use interactive mode for get a command line inside the container:

Master_Header.sh calls two separate header scripts:

1. Module_Search_XXXXX.sh, which will look for vdj 10-mers in bam files. By default this looks for bams in the root of the mounted folder. note that Module_Search_XXXXX.sh will create multiple bam files and index files for the individual receptors. Module_Search_XXXXX.sh can be edited and a different path in the mounted folder can be set.
2. VDJ_header.sh, which does pairwise alignment (against known VDJ sequences in /vdj_recovery/shared/db in the container). VDJ_header.sh can be edited to specify a different path, to only search for certain receptors, and to specify a samples file to specify a samples file add it to this line in VDJ_header.sh: python " /vdj_recovery/shared/vdjrecord_header.py" '/mnt/results/' example: python " /vdj_recovery/shared/vdjrecord_header.py" '/mnt/results/' '/mnt/samples.csv' if the samples.csv file is in the root of the mounted volume

Samples file: This file can be used if the filenames of the .bam files are not directly representative of case IDs. For example, if the filename is something like 1234.bam, and that name represents a specific case ID e.g. "CASE01" and a specific sample e.g. "Tumor Tissue" If you want the script to automatically match the .bam filename to that Case ID and Sample, create a .csv in the mounted directory with the following columns:

An example samples file can be found in the container at /vdj_recovery/shared/example_samples.csv