Parameters
Set the required inputs and the parameters that most often change a multiomics workflow run.
Most runs need only the run directory. Everything else has a default, including the reference genome. This page covers the parameters that matter most; the complete reference with every parameter and its default is docs/parameters.md in the workflow repository.
Pass parameters on the command line (--run_dir ...) or in a params file (-params-file params.json). Lists such as tile need the params file.
Required
| Parameter | Description |
|---|---|
run_dir | The AVITI24™ cell screening run directory, local path or S3 URI. Must contain RunParameters.json, RunManifest.csv, and Panel.json. |
Reference selection
| Parameter | Default | Description |
|---|---|---|
genome | GRCh38 | Reference genome. Selects the FASTA, a STAR index built with STAR 2.7, and the GENCODE release 47 annotation. GRCh38 is the only supported genome. |
igenomes_base | Element Biosciences reference bucket | Base path of the reference layout. Point it at a copy of that layout to avoid remote reads. |
genes_gtf | from genome | Replaces the genome's GENCODE annotation with another GENCODE release on the same assembly. Release 47 is validated; releases from 31 on are expected to run but are not validated. |
igenomes_ignore | false | Use a custom genome instead: set true and pass fasta, star_index, and genes_gtf together. |
fasta, star_index | from genome | Custom genome FASTA and STAR index, built with STAR 2.7.4a or later from that FASTA. Accepted only with igenomes_ignore. The matching genes_gtf must be GENCODE-format, which in practice means human or mouse. |
Scope of a run
| Parameter | Default | Description |
|---|---|---|
id | detected from run_dir | Run identifier used in output names. |
outdir | ./results | Output directory, local or S3. |
tile | all tiles | List of tile ID patterns, for example ["L1R01C01S1"]. The same subset applies to Cells2Stats, Bases2Fastq™, and the SDO build. |
filter_batch | all batches | Comma-delimited batch list, for example B01,B02. |
segmentation_dir | from the run | Pre-computed cell segmentation output, when Cells2Stats ran separately. |
3′ counting options
These parameters change the count matrix; leave them at their defaults unless you are comparing counting policies.
| Parameter | Default | Description |
|---|---|---|
f2p_3p_features | utr,exon,chrm | Feature tiers that count: 3′ UTR tiers, the exon union (catches UTR-less genes), and the chrM umbrella. Add intron to count intronic reads. |
f2p_3p_include_gene_types | protein_coding,lncRNA | Gene biotypes retained when building the counting features. |
f2p_3p_count_simple_multimapping_reads | true | Count reads that multimap within a single gene. |
f2p_3p_unique_only_tiers | intron | Tiers counted unique-only (their simple multimappers dropped); inert unless the tier is in f2p_3p_features. |
f2p_3p_slop_3p_bp | 100 | Extension past the gene's 3′ terminus, in bp. |
f2p_3p_slop_5p_bp | 200 | Extension in the 5′ direction into the CDS, in bp. |
Reports and outputs
| Parameter | Default | Description |
|---|---|---|
enable_c2s_qc_report | true | Generate the Cells2Stats QC report (cells2stats/qc/qc_report.html). |
avsd_qc_enabled | false | Generate the optional run-quality report. Requires spatialdata_analysis_container_url / _tag; the report image is not part of the public release. |
avsd_zip_spatialdata | true | Zip the .zarr store into {run_id}.zarr.zip. When false, only the directory is written. |
Container images
Each release defaults to the image tags it was validated with. Override one or more with c2s_container_url / c2s_container_tag, b2f_container_url / b2f_container_tag, elembiodisstools_container_url / elembiodisstools_container_tag, and spatialdata_build_container_url / spatialdata_build_container_tag. Spell Docker Hub images with the registry (docker.io/elembio/...).