A Nextflow script that will run AMRFinderPlus on fasta files on S3 and databases available on a shared filesystem (EFS).
aws batch submit-job \
--job-name nf-amr-TEST \
--job-queue priority-maf-pipelines \
--job-definition nextflow-production \
--container-overrides command=fischbachLab/nf-amr,\
"--project","00_TEST",\
"--prefix","20230822",\
"--seedfile","s3://nextflow-pipelines/nf-amr/data/sample_seedfile.csv"The seedfile is a CSV file with the following columns: sample_name,fasta_file. A sample seedfile is available in data/sample_seedfile.csv.
As of 2023/09/09, the date of implementing this pipeline, there is an error in AMRFinderPlus that has been difficult to reproduce reliably. This error causes the application to exit with Segmentation Fault apparently on random genomes. It has also been discussed on the official AMRFinderPlus repo as an issue here.
In order to work around it, we simply retry the job upto 5 times and that seems to work. Until a fix is available, this is the best we can do.