Native Backends

CBIcall separates what analysis is run from how it is executed. The analysis is selected with pipeline and mode; the execution technology is selected with workflow_backend.

A pipeline implementation is CBIcall-native when it produces the CBIcall output contract: the generated cbicall_* run directory, standard logs, audit reports, output inventory, and final-output fingerprints when available. Native status is about that output contract, not the workflow language.

Supported Native Backends

Backend	Role	Workflow form	Required runner
Bash	Default and recommended	Shell scripts	Bash
Snakemake	Illustrative and backend-equivalence testing	Snakefile rules	Snakemake
Nextflow	Illustrative and backend-equivalence testing	Nextflow processes	Nextflow and Java
Cromwell	Illustrative and backend-equivalence testing	WDL workflows and tasks	Cromwell and Java

Recommended backend

Bash is the default and recommended backend for routine use of the bundled cbicall-core pipelines. The Snakemake, Nextflow, and Cromwell implementations demonstrate backend portability and support backend-equivalence testing.

What cbicall-core means

CBIcall is the framework. cbicall-core is the ready-to-run workflow collection distributed and maintained with it.

External nf-core provider entries are different: they are launched through the Nextflow backend, but they keep their upstream nf-core output layout and runtime assumptions. See External nf-core.

Use cbicall doctor to see which backend runners are available in the current installation.

Selecting a Backend

pipeline: wes
mode: single
workflow_provider: cbicall-core
workflow_backend: bash
software_stack: gatk-4.6

For these bundled workflows, workflow_provider: cbicall-core is the default and can be omitted. Keep workflow_backend: bash for routine use. The other backends can be selected when testing portability or backend equivalence.

Use Included Pipelines for the compatibility matrix and pipeline-specific guides.