Pathogenome AMR workflow

This workflow is currently a beta release.

This Nextflow workflow starts with a list of NCBI or ENA sequencing-run accessions and looks for antimicrobial resistance (AMR) genes in their assembled genomes. It is configured to run on the Seqera Platform and uses the pathogenome distribution of QIIME 2 under the hood.

What the workflow does

1. Imports accession IDs.
2. Downloads the sequencing data with q2-fondue.
3. Assembles paired-end reads into contigs with MEGAHIT.
4. Predicts genes and proteins with Prodigal.
5. Downloads the AMRFinderPlus database and annotates the predicted genes and proteins for AMR-related features.

Only paired-end reads are passed to the assembly step.

Required inputs

Accession file

Provide a tab-separated text file with a sample-id header and one NCBI or ENA run accession per row:

sample-id
ERR00000000
ERR00000001

Upload this file to a location available in Seqera Data Explorer, or add it as a Seqera dataset.

Contact email

The q2-fondue step requires an email address (because NCBI requires one). Supply one on the launch page on Seqera.

Launching on Seqera

1. Open Launchpad in your Seqera workspace and add or select this pipeline.
2. Choose the latest beta tag. To test unreleased changes from main, select main and enable Pull latest.
3. Select your Google Batch compute environment and Google Cloud Storage work directory.
4. Leave Config profile empty and set Pipeline schema to Repository default.
5. In Run parameters, select the accession TSV for Accessions and enter your address for Fondue email.
6. Optionally adjust the CPU/thread, memory, and time limits for each step. Memory is specified in GB and time limits are specified in hours.
7. Select Launch.

Default resources

Step	CPUs/threads	Memory	Time limit
Accession import	1	8 GB	1 hour
Fondue download	8	16 GB	36 hours
MEGAHIT assembly	12	64 GB	48 hours
Prodigal gene prediction	12	32 GB	24 hours
AMRFinderPlus database download	8	16 GB	24 hours
AMR annotation	4	128 GB	24 hours

Fondue, MEGAHIT, and AMRFinderPlus receive their values as application thread counts. For the other steps, the CPU value controls the compute resources requested from Seqera, although the underlying application may not use every allocated CPU.

Finding the results

After the run finishes, open Tasks, select the AMRAnnotate task, and open the Data Explorer tab. The QIIME 2 outputs are:

• amrfinderplus_annotations.qza
• amrfinderplus_all_mutations.qza
• amrfinderplus_genes.qza
• amrfinderplus_proteins.qza

Intermediate artifacts such as the reads downloaded with q2-fondue, or the contigs from assembly can be found in the Data Explorer tabs of the corresponding tasks.

Troubleshooting a failed run

If the workflow fails, open the run in Seqera, select Tasks, select the failed task, and then select its Execution log tab. The execution log will contain information about the error that occurred. For help interpreting the error, please don't hesitate to reach out on the QIIME 2 Forum.