Process packs¶
A process pack is the omics modality plus the catalog rows that specialize it. GoliathOmics owns the JSON and the analyte SQL. GoliathApp owns the cfg tables those rows land in. Sync is scripts/sync_cfg_profiles_and_action_catalog.py --backend postgres, usually via scripts/deploy_process_pack_catalog.sh --backend postgres --sync. See Release.
Three layers¶
| Layer | Operator field | Where it lives | What it selects |
|---|---|---|---|
| Modality | regulatory.primary_modality |
Profile and study. Enforced by workflow_engine/domain/modality_gate.py |
methylation, rnaseq, or proteomics |
| Analyte | regulatory.primary_analyte |
workflow_engine/domain/analytes/*.analyte.json → cfg.analyte |
Sample matrix |
| Technique | libraryProtocol on an assay procedure |
workflow_engine/domain/profiles/procedures/*.procedure.json → cfg.assay_procedure |
How the library was made |
| Disease or trait | diseases / controls on the project manifest, plus pipelineProcedure |
schemas/config/project_config.schema.json and docs/examples/samd/ |
The study cohort on top of a procedure |
There is no cfg.disease catalog in this repository. A disease or trait study is an application pack: a project manifest and a small actionConfig overlay on a procedure that already exists. Chapters 21, 23, and 25 are those studies. Chapter 24 is the pattern.
Analytes¶
name |
File |
|---|---|
buffy_coat |
workflow_engine/domain/analytes/buffy_coat.analyte.json |
cfdna |
workflow_engine/domain/analytes/cfdna.analyte.json |
tissue |
workflow_engine/domain/analytes/tissue.analyte.json |
plant_tissue |
workflow_engine/domain/analytes/plant_tissue.analyte.json |
combined |
workflow_engine/domain/analytes/combined.analyte.json |
The study primary_analyte must match the procedure's analyteExpectation. Fill-in behavior for QC and deconvolution is in Analyte profiles.
Techniques¶
schemas/config/procedure.schema.json allows libraryProtocol values wgbs_linear, wgbs_pangenome, epi_gbs, and emseq_targeted.
| Procedure | Analyte | libraryProtocol |
|---|---|---|
buffy_wgbs_pangenome_gene_fc |
buffy_coat |
wgbs_pangenome |
buffy_wgbs_linear_gene_fc |
buffy_coat |
wgbs_linear |
buffy_wgbs_linear_mojo_gene_fc |
buffy_coat |
wgbs_linear (parabricks.engine mojo) |
buffy_wgbs_mvalue_residual_gene_fc |
buffy_coat |
wgbs_pangenome |
cfdna_wgbs_plasma |
cfdna |
wgbs_linear |
cfdna_emseq_targeted |
cfdna |
emseq_targeted |
cfdna_emseq_mhl_survival |
cfdna |
emseq_targeted |
plant_wgbs_gene_fc |
plant_tissue |
wgbs_linear |
Modality programs¶
| Modality | SamplePrep program | Lifecycle program | Profile |
|---|---|---|---|
| Methylation | sample_prep.program.json (and sample_prep_emseq, sample_prep_epigbs, sample_prep_remediate) |
study_validation_lifecycle.program.json |
samd_research and the other methylation profiles under workflow_engine/domain/profiles/ |
| RNA-seq | sample_prep_rnaseq.program.json |
rnaseq_study_lifecycle.program.json |
rnaseq_research |
| Proteomics | sample_prep_proteomics.program.json |
proteomics_study_lifecycle.program.json |
proteomics_research |
scripts/deploy_workflow_definitions.sh compiles the methylation trio always, and the RNA-seq and proteomics programs when the fixture files are present. Chapter 20 and chapter 22 describe those two modalities.