CLI Workflows
What this page is for
Use this page when you want to run MicroSeq from a terminal, inside a bash wrapper, over SSH, or on an HPC cluster. It documents the current headless paired-AB1 workflow and shows how the wrapper maps onto the underlying MicroSeq CLI stages.
This page is operational. For pairing rules, assembly semantics, ambiguity handling, and BLAST-input behavior, see the Paired CAP3 Assembly Tutorial. For the meaning of output files, see Output Artifacts Reference.
When to use this page instead of the GUI
Use the CLI/HPC workflow when:
- you want reproducible batch runs
- you are working on a remote server or cluster
- you want to call MicroSeq from a shell script or scheduler
- you want one run-scoped log across multiple stages
- you want a stable operational workflow that mirrors the paired pipeline without relying on interactive dialogs
MicroSeq CLI / HPC Quickstart
Install MicroSeq
For installation, environment setup, and database setup, follow the main README:
- GitHub repo: https://github.com/jose-cantu/MicroSeq
- README: https://github.com/jose-cantu/MicroSeq/blob/master/README.MD
This includes:
- creating and activating the
MicroSeqconda environment - installing the package
- running
microseq-setup - configuring databases and logging
Once installation is complete, activate the environment:
conda activate MicroSeq
Current paired AB1 HPC workflow
For using the bash wrappers I have setup refer to this readme I made here:
- AB1 CLI README: https://github.com/jose-cantu/MicroSeq/tree/master/workflows
The current bash wrapper for paired AB1 runs is:
bash workflows/paired_ab1_pipeline.sh <input_dir> <db_key> <blast_threads> [out_dir]
Arguments:
input_dir= directory containing raw paired.ab1filesdb_key=gg2,silva, orncbiblast_threads= number of threads to pass to BLASTout_dir= optional output directory; default is<input_dir>_microseq
Minimal example
conda activate MicroSeq
bash workflows/paired_ab1_pipeline.sh \
tests/paired_single_pair_ab1_demo_run \
gg2 \
4
Optional primer-label override
If your forward and reverse primer labels differ from the demo defaults, override them before running the wrapper:
conda activate MicroSeq
export FWD_PATTERN="27F"
export REV_PATTERN="1492R"
bash workflows/paired_ab1_pipeline.sh \
/path/to/paired_ab1_inputs \
gg2 \
4
Optional run-scoped logging
If you want every microseq stage launched by the wrapper to write into one run-scoped log context:
export MICROSEQ_SESSION_ID="paired-run-001"
Then run the wrapper normally.
What the wrapper currently does
The wrapper currently runs the paired AB1 workflow in this order:
- trim AB1 -> QC-passed FASTQ
- stage paired FASTA/QUAL
- write
qc/pairing_report.tsv - run paired assembly
- write
asm/assembly_summary.tsv - write
qc/overlap_audit.tsv - write
asm/blast_inputs.fastaandasm/blast_inputs.tsv - run BLAST
- join taxonomy
Key outputs
After a successful run, the main outputs are:
qc/trim_summary.tsvqc/pairing_report.tsvqc/overlap_audit.tsvasm/assembly_summary.tsvasm/blast_inputs.fastaasm/blast_inputs.tsvhits.tsvhits_tax.tsv
Notes
- Use the README for installation and setup.
- Use the 2nd README in
/workflowfor the preview pairs wrapper and this wrapper for the current headless paired AB1 workflow. - For pairing/assembly behavior and output semantics, see the paired assembly documentation in
/docs.