Parameters¶
Complete reference for all --param flags accepted by main.nf. Defaults are set in configs/base.config. The machine-readable version of this reference is nextflow_schema.json.
Input / output options¶
| Parameter | Default | Description |
|---|---|---|
--workflow |
integration |
Named workflow to execute. One of integration, ingest_export, ingest_tabulate, nmf_vae, gex_mil, tcr_mil, tcr_epitope, make_tcr_vector_database. |
--input |
data/samplesheet.csv |
Path to the samplesheet CSV. See Data Formats → Samplesheet. |
--outdir |
outputs/ |
Directory where published results are written. |
LabKey / Prime-seq options¶
These are only required when at least one row in the samplesheet uses output_file_id (LabKey mode). Rows using url or path do not need LabKey credentials.
| Parameter | Default | Description |
|---|---|---|
--labkey_base_url |
(required for LabKey rows) | Base URL of the LabKey server (e.g. https://labkey.example.org). |
--labkey_folder |
(required for LabKey rows) | LabKey folder path (e.g. /My/Project/Folder). |
Species options¶
| Parameter | Default | Description |
|---|---|---|
--species_order |
human,macaque,mouse |
Comma-separated list of species. Controls the ordering during harmonization and scMODAL integration. Each value must match the species column in the samplesheet. |
--export_assay |
RNA |
Seurat assay name to export as count matrix in EXPORT_COUNTS. |
Tabulation options¶
These parameters affect only --workflow ingest_tabulate.
| Parameter | Default | Description |
|---|---|---|
--tabulate_id_cols |
cDNA_ID,SubjectId,Vaccine,Timepoint,Tissue |
Comma-separated subject-level identity columns to carry into subjectIdTable.csv. cDNA_ID is always included as the primary sample key. |
--tabulate_celltype_cols |
(empty) | Extra cell-type annotation columns to tabulate beyond the standard RIRA set. Leave empty to use only the standard RIRA columns auto-detected in the metadata. |
--tabulate_parent_col |
(empty) | Parent lineage column used to gate child cell-type columns. Defaults to RIRA_Immune.cellclass when empty. |
--tabulate_celltype_parent_map |
(empty) | Comma-separated celltype_col:parentValue pairs to extend or override the built-in hierarchy. Example: RIRA_TNK_v2.cellclass:TNK,RIRA_Myeloid_v3.cellclass:Myeloid. |
scMODAL integration options¶
These parameters affect only --workflow integration.
| Parameter | Default | Description |
|---|---|---|
--scmodal_container |
ghcr.io/gwmcelfresh/scmodal:latest |
Container image for GENE_HARMONIZE and SCMODAL_INTEGRATE. Must include scmodal, torch, scanpy, and anndata. |
--scmodal_latent |
20 |
Number of latent dimensions in the scMODAL VAE embedding. |
--scmodal_training_steps |
10000 |
Number of VAE training steps. Increase for larger datasets. |
--scmodal_batch_size |
500 |
Mini-batch size during scMODAL training. |
--scmodal_neighbors |
30 |
Number of nearest neighbours for the KNN graph built on the latent embedding. |
--leiden_resolution |
0.5 |
Leiden clustering resolution. Higher values produce more clusters. |
CI / testing options¶
Not for production use
--scmodal_use_cpu is intended exclusively for GitHub Actions smoke tests.
Using it outside CI will produce stub outputs with no scientific validity and will print a warning.
| Parameter | Default | Description |
|---|---|---|
--scmodal_use_cpu |
false |
Bypasses the local-executor GPU guard for --workflow integration and runs SCMODAL_INTEGRATE as a stub (requires -stub-run). Emits a warning if GITHUB_ACTIONS env is not set. |
Generic options¶
| Parameter | Default | Description |
|---|---|---|
--help |
false |
Print help text and exit. |
HPC execution environment¶
The following environment variables control how rootless Podman manages container images and scratch storage on HPC clusters. They are set in the shell before launch, not passed as --param flags.
See Section 7 of the usage guide for the full HPC setup walkthrough.
| Variable | Description |
|---|---|
NXF_PODMAN_GRAPHROOT |
Path to your existing shared Podman image store (NFS-backed). Auto-detected from podman info; override if auto-detection is unavailable or points to a quota-constrained home directory. |
NXF_PODMAN_LOCAL_SCRATCH |
Optional path override for the node-local scratch base used for Podman runroot, TMPDIR, and XDG_RUNTIME_DIR. Image layers stay in NXF_PODMAN_GRAPHROOT. |
NXF_PODMAN_PULL_LOCK_DIR |
Shared lock directory to serialize concurrent image pulls. Defaults to ${NXF_WORK}/.podman-pull-locks. |