Run controls through cttso pipeline

[alexiswl]

Under dragen_tso_ctdna_step.py snippet below; we currently remove samples that have qc selected and only include samples that must be tumor.

data-portal-apis/data_processors/pipeline/orchestration/dragen_tso_ctdna_step.py

Lines 91 to 97 in f790a68

	LabMetadataRule(this_metadata) \
	.must_set_workflow() \
	.must_not_manual() \
	.must_not_bcl() \
	.must_not_qc() \
	.must_be_cttso_ctdna() \
	.must_be_tumor()

If we wish to also run the control samples, should this rule be reduced to the following criteria?

must_set_workflow()
must_not_bcl()
must_be_cttso_ctdna()

[victorskl]

Ok. I reckon, we still need not being manual flag set i.e. it evaluates Lab meta workflow column:

data-portal-apis/data_processors/pipeline/domain/workflow.py

Lines 323 to 327 in f790a68

	def must_not_manual(self):
	from data_portal.models.labmetadata import LabMetadataWorkflow
	if self.this_metadata.workflow.lower() == LabMetadataWorkflow.MANUAL.value.lower():
	raise LabMetadataRuleError(f"Workflow is set to manual.")
	return self

[victorskl]

Think, tumor flag is required, purely based on workflow type name there.. i.e. tso_ctdna_tumor_only. So, any sample flag not tumor get better filtered out.

data-portal-apis/data_processors/pipeline/domain/workflow.py

Lines 349 to 354 in f790a68

	def must_be_tumor(self):
	"""Sample Phenotype must be tumor"""
	from data_portal.models.labmetadata import LabMetadataPhenotype
	if self.this_metadata.phenotype.lower() != LabMetadataPhenotype.TUMOR.value.lower():
	raise LabMetadataRuleError(f"Not tumor sample.")
	return self

data-portal-apis/data_processors/pipeline/domain/workflow.py

Lines 30 to 37 in f790a68

	class WorkflowType(Enum):
	BCL_CONVERT = "bcl_convert"
	DRAGEN_WGS_QC = "wgs_alignment_qc"
	TUMOR_NORMAL = "wgs_tumor_normal"
	DRAGEN_TSO_CTDNA = "tso_ctdna_tumor_only"
	DRAGEN_WTS = "wts_tumor_only"
	UMCCRISE = "umccrise"
	RNASUM = "rnasum"

[victorskl]

Historically, qc flag use to annotate; say that those samples that wish only want FASTQs and, run alignment QC if WGS.

data-portal-apis/data_processors/pipeline/domain/workflow.py

Lines 329 to 334 in f790a68

	def must_not_qc(self):
	"""Demultiplex and run QC on FASTQ (once on ICA that includes DRAGEN alignment for better stats)"""
	from data_portal.models.labmetadata import LabMetadataWorkflow
	if self.this_metadata.workflow.lower() == LabMetadataWorkflow.QC.value.lower():
	raise LabMetadataRuleError(f"Workflow is set to QC.")
	return self

[victorskl]

I guess; without changing to current LabMetadataRule and as-is; in order to run TSO control samples through ctTSO pipeline, we annotate them in Lab tracking sheet as follows.

Type: ctDNA
Assay: ctTSO
Workflow: control
Phenotype: tumor

That should work?

@mhlunimelb @alexiswl @ohofmann @reisingerf

[victorskl]

Descope from 2.0 release for now further discussion