fix: isa-tab add-ped now only modifies values on process nodes of ped-supplied samples #207

cubi_tk/isa_tab/add_ped.py

@@ -178,6 +178,13 @@ def on_visit_material(self, material, node_path, study=None, assay=None):
 
     def on_visit_process(self, process, node_path, study=None, assay=None):
         super().on_visit_process(process, node_path, study, assay)
+        donor_name = (
+            node_path[0].name
+            if _is_source(node_path[0])
+            else _sample_to_donor_name(node_path[0].name)
+        )
+        if donor_name not in self.donor_map:
+            return None
         proc_config_pairs = {
             "library construction ": {
                 "library type": "library_type",
@@ -479,6 +486,10 @@ def _append_assay_line_protocol(
     return counter, curr
 
 
+def _sample_to_donor_name(sample_name):
+    return sample_name.rstrip("-N1")
+
+
 def _donor_to_sample_name(donor_name):
     return "%s-N1" % donor_name
 

tests/data/isa_tab/expected_output2/a_test_isa_tpl_add_ped_exome_sequencing_nucleotide_sequencing.txt

@@ -0,0 +1,4 @@
+Sample Name	Protocol REF	Parameter Value[Concentration measurement]	Performer	Date	Extract Name	Characteristics[Concentration]	Unit	Term Source REF	Term Accession Number	Protocol REF	Parameter Value[Provider name]	Parameter Value[Provider contact]	Parameter Value[Provider project ID]	Parameter Value[Provider sample ID]	Parameter Value[Provider QC status]	Parameter Value[Requestor contact]	Parameter Value[Requestor project]	Parameter Value[Requestor sample ID]	Parameter Value[Concentration measurement]	Parameter Value[Library source]	Parameter Value[Library strategy]	Parameter Value[Library selection]	Parameter Value[Library layout]	Parameter Value[Library kit]	Comment[Library kit catalogue ID]	Parameter Value[Target insert size]	Parameter Value[Wet-lab insert size]	Parameter Value[Barcode kit]	Parameter Value[Barcode kit catalogue ID]	Parameter Value[Barcode name]	Parameter Value[Barcode sequence]	Performer	Date	Library Name	Characteristics[Folder name]	Characteristics[Concentration]	Unit	Term Source REF	Term Accession Number	Protocol REF	Parameter Value[Platform]	Parameter Value[Instrument model]	Parameter Value[Base quality encoding]	Parameter Value[Center name]	Parameter Value[Center contact]	Performer	Date	Raw Data File
+index-N1	Nucleic acid extraction WES				index-N1-DNA1					Library construction WES										GENOMIC	WXS	Hybrid Selection	PAIRED	Agilent SureSelect Human All Exon V6r2	S00000XYZ									index-N1-DNA1-WES1	index					Nucleic acid sequencing WES	ILLUMINA	Illumina NextSeq 500	Phred+33					
+father-N1	Nucleic acid extraction WES				father-N1-DNA1					Library construction WES										GENOMIC	WXS	Hybrid Selection	PAIRED											father-N1-DNA1-WES1	father					Nucleic acid sequencing WES			Phred+33					
+mother-N1	Nucleic acid extraction WES				mother-N1-DNA1					Library construction WES										GENOMIC	WXS	Hybrid Selection	PAIRED											mother-N1-DNA1-WES1	mother					Nucleic acid sequencing WES			Phred+33					

tests/data/isa_tab/expected_output2/i_Investigation.txt

@@ -0,0 +1,95 @@
+ONTOLOGY SOURCE REFERENCE
+Term Source Name	UO	OBI	NCBITAXON	OMIM	HP	ORPHA
+Term Source File	http://data.bioontology.org/ontologies/UO	http://data.bioontology.org/ontologies/OBI	http://data.bioontology.org/ontologies/NCBITAXON	http://data.bioontology.org/ontologies/OMIM	http://data.bioontology.org/ontologies/HP	http://data.bioontology.org/ontologies/ORDO
+Term Source Version	48	35	10	12	570	2.8
+Term Source Description	Units of Measurement Ontology	Ontology for Biomedical Investigations	National Center for Biotechnology Information (NCBI) Organismal Classification	Online Mendelian Inheritance in Man	Human Phenotype Ontology	Orphanet Rare Disease Ontology
+INVESTIGATION
+Investigation Identifier	
+Investigation Title	Test ISA-Tab add-ped
+Investigation Description	
+Investigation Submission Date	
+Investigation Public Release Date	
+Comment[Created With Configuration]	/path/to/isa-configurations/bih_studies/bih_germline
+Comment[Last Opened With Configuration]	bih_germline
+INVESTIGATION PUBLICATIONS
+Investigation PubMed ID
+Investigation Publication DOI
+Investigation Publication Author List
+Investigation Publication Title
+Investigation Publication Status
+Investigation Publication Status Term Accession Number
+Investigation Publication Status Term Source REF
+INVESTIGATION CONTACTS
+Investigation Person Last Name
+Investigation Person First Name
+Investigation Person Mid Initials
+Investigation Person Email
+Investigation Person Phone
+Investigation Person Fax
+Investigation Person Address
+Investigation Person Affiliation
+Investigation Person Roles
+Investigation Person Roles Term Accession Number
+Investigation Person Roles Term Source REF
+STUDY
+Study Identifier	test_isa_tpl_add_ped
+Study Title	Test ISA-Tab add-ped
+Study Description	
+Comment[Study Grant Number]	
+Comment[Study Funding Agency]	
+Study Submission Date	
+Study Public Release Date	
+Study File Name	s_test_isa_tpl_add_ped.txt
+STUDY DESIGN DESCRIPTORS
+Study Design Type
+Study Design Type Term Accession Number
+Study Design Type Term Source REF
+STUDY PUBLICATIONS
+Study PubMed ID
+Study Publication DOI
+Study Publication Author List
+Study Publication Title
+Study Publication Status
+Study Publication Status Term Accession Number
+Study Publication Status Term Source REF
+STUDY FACTORS
+Study Factor Name
+Study Factor Type
+Study Factor Type Term Accession Number
+Study Factor Type Term Source REF
+STUDY ASSAYS
+Study Assay File Name	a_test_isa_tpl_add_ped_exome_sequencing_nucleotide_sequencing.txt
+Study Assay Measurement Type	exome sequencing
+Study Assay Measurement Type Term Accession Number	
+Study Assay Measurement Type Term Source REF	
+Study Assay Technology Type	nucleotide sequencing
+Study Assay Technology Type Term Accession Number	http://purl.obolibrary.org/obo/OBI_0000626
+Study Assay Technology Type Term Source REF	OBI
+Study Assay Technology Platform	Illumina
+STUDY PROTOCOLS
+Study Protocol Name	Sample collection	Nucleic acid extraction WES	Library construction WES	Nucleic acid sequencing WES
+Study Protocol Type	Sample collection	Nucleic acid extraction WES	Library construction WES	Nucleic acid sequencing WES
+Study Protocol Type Term Accession Number				
+Study Protocol Type Term Source REF				
+Study Protocol Description				
+Study Protocol URI				
+Study Protocol Version				
+Study Protocol Parameters Name		Concentration measurement	Provider project ID;Library source;Library selection;Library layout;Barcode sequence;Wet-lab insert size;Requestor contact;Library kit;Barcode name;Provider contact;Concentration measurement;Provider sample ID;Target insert size;Provider name;Requestor project;Requestor sample ID;Barcode kit;Barcode kit catalogue ID;Provider QC status;Library strategy	Base quality encoding;Platform;Center contact;Instrument model;Center name
+Study Protocol Parameters Name Term Accession Number			;;;;;;;;;;;;;;;;;;;	;;;;
+Study Protocol Parameters Name Term Source REF			;;;;;;;;;;;;;;;;;;;	;;;;
+Study Protocol Components Name				
+Study Protocol Components Type				
+Study Protocol Components Type Term Accession Number				
+Study Protocol Components Type Term Source REF				
+STUDY CONTACTS
+Study Person Last Name
+Study Person First Name
+Study Person Mid Initials
+Study Person Email
+Study Person Phone
+Study Person Fax
+Study Person Address
+Study Person Affiliation
+Study Person Roles
+Study Person Roles Term Accession Number
+Study Person Roles Term Source REF

tests/data/isa_tab/expected_output2/s_test_isa_tpl_add_ped.txt

@@ -0,0 +1,4 @@
+Source Name	Characteristics[UUID]	Characteristics[External links]	Characteristics[Batch]	Characteristics[Family]	Characteristics[Organism]	Term Source REF	Term Accession Number	Characteristics[Mother]	Characteristics[Father]	Comment[Family notes]	Characteristics[Sex]	Characteristics[Disease status]	Characteristics[OMIM disease]	Term Source REF	Term Accession Number	Characteristics[Orphanet disease]	Term Source REF	Term Accession Number	Characteristics[HPO terms]	Term Source REF	Term Accession Number	Comment[Disease notes]	Protocol REF	Performer	Sample Name	Characteristics[External links]	Characteristics[Cell origin]	Term Source REF	Term Accession Number	Characteristics[Cell type]	Term Source REF	Term Accession Number
+index		x-charite-medgen-blood-book-id:index	3	FAM_index	Homo sapiens	NCBITAXON	http://purl.bioontology.org/ontology/NCBITAXON/9606	mother	father		female	affected											Sample collection		index-N1							
+father		x-charite-medgen-blood-book-id:father	.	FAM_index	Homo sapiens	NCBITAXON	http://purl.bioontology.org/ontology/NCBITAXON/9606	0	0		male	unaffected											Sample collection		father-N1							
+mother		x-charite-medgen-blood-book-id:mother	.	FAM_index	Homo sapiens	NCBITAXON	http://purl.bioontology.org/ontology/NCBITAXON/9606	0	0		female	unaffected											Sample collection		mother-N1							

tests/data/isa_tab/in_sheet_subset/input.ped

@@ -0,0 +1 @@
+FAM	index	father	mother	2	2

tests/test_isa_tab_add_ped.py

@@ -100,3 +100,28 @@ def test_add_ped_just_update(tmpdir):
         str(scratch_dir),
         str(pathlib.Path(__file__).parent / "data" / "isa_tab" / "expected_output"),
     )
+
+
+def test_add_ped_sheet_subset_update(tmpdir):
+    """Test updating study and assay, but only a subset of the sheet."""
+    scratch_dir = tmpdir / "scratch"
+    path_ped = pathlib.Path(__file__).parent / "data" / "isa_tab" / "in_sheet_subset" / "input.ped"
+    shutil.copytree(
+        str(pathlib.Path(__file__).parent / "data" / "isa_tab" / "in_just_update"), str(scratch_dir)
+    )
+
+    # Update metadata
+    args = BASE_ARGS[:]
+    args[12] = "S00000XYZ"
+
+    argv = args + [str(scratch_dir / "i_Investigation.txt"), str(path_ped)]
+
+    # Actually exercise code and perform test.
+    res = main(argv)
+
+    assert not res
+
+    compare_input_output(
+        str(scratch_dir),
+        str(pathlib.Path(__file__).parent / "data" / "isa_tab" / "expected_output2"),
+    )