From 5c05facbdb1e35556256cd11d6a814822ebb6159 Mon Sep 17 00:00:00 2001
From: lltommy
Date: Thu, 30 Apr 2020 18:00:58 +0200
Subject: Last addition/edits to the meta data schema

---
 bh20sequploader/bh20seq-schema.yml    | 15 ++++++++++++---
 bh20sequploader/bh20seq-shex.rdf      |  5 +++--
 example/maximum_metadata_example.yaml |  1 +
 example/minimal_metadata_example.yaml |  2 +-
 4 files changed, 17 insertions(+), 6 deletions(-)

diff --git a/bh20sequploader/bh20seq-schema.yml b/bh20sequploader/bh20seq-schema.yml
index d3e992a..99e1a11 100644
--- a/bh20sequploader/bh20seq-schema.yml
+++ b/bh20sequploader/bh20seq-schema.yml
@@ -49,6 +49,7 @@ $graph:
         jsonldPredicate:
           _id: http://purl.obolibrary.org/obo/NCIT_C25688
           _type: "@id"
+          noLinkCheck: true
     host_treatment:
       doc: Process in which the act is intended to modify or alter host status
       type: string?
@@ -59,6 +60,13 @@ $graph:
       type: string[]?
       jsonldPredicate:
           _id: http://purl.obolibrary.org/obo/VO_0000002
+    ethnicity:
+      doc: Ethinicity of the host e.g. http://purl.obolibrary.org/obo/HANCESTRO_0010
+      type: string?
+      jsonldPredicate:
+          _id: http://semanticscience.org/resource/SIO_001014
+          _type: "@id"
+          noLinkCheck: true
     additional_host_information:
       doc: Field for additional host information
       type: string?
@@ -91,7 +99,7 @@ $graph:
       jsonldPredicate:
           _id: http://purl.obolibrary.org/obo/OBI_0001895
     collecting_institution:
-      doc: Institute that was responsible of sampeling
+      doc: Institute that was responsible for sampeling
       type: string?
       jsonldPredicate:
           _id: http://purl.obolibrary.org/obo/NCIT_C41206
@@ -118,12 +126,13 @@ $graph:
       jsonldPredicate:
           _id: http://edamontology.org/data_2091
           _type: "@id"
+          noLinkCheck: true
 
 - name: virusSchema
   type: record
   fields:
     virus_species:
-      doc: The name of a taxon from the NCBI taxonomy database
+      doc: The name of virus species from the NCBI taxonomy database, e.g. http://purl.obolibrary.org/obo/NCBITaxon_2697049 for Severe acute respiratory syndrome coronavirus 2
       type: string
       jsonldPredicate:
           _id: http://edamontology.org/data_1875
@@ -152,7 +161,7 @@ $graph:
         _id: http://www.ebi.ac.uk/efo/EFO_0002699
     sequencing_coverage:
       doc: Sequence coverage defined as the average number of reads representing a given nucleotide (e.g. [100]) - if multiple technologies were used multiple float values can be submitted e.g. [100, 20]
-      type: float[]?
+      type: double[]?
       jsonldPredicate:
         _id: http://purl.obolibrary.org/obo/FLU_0000848
     additional_technology_information:
diff --git a/bh20sequploader/bh20seq-shex.rdf b/bh20sequploader/bh20seq-shex.rdf
index fe6deee..cdf2296 100644
--- a/bh20sequploader/bh20seq-shex.rdf
+++ b/bh20sequploader/bh20seq-shex.rdf
@@ -27,6 +27,7 @@ PREFIX wikidata: <http://www.wikidata.org/entity/>
     efo:EFO_0000727 xsd:string ?;
     obo:VO_0000002 xsd:string {0,10};
     sio:SIO_001167 xsd:string ?;
+    sio:SIO_001014 [ obo:HANCESTRO_~ ] ? ; #ethnicity
 }
 
 :sampleShape  {
@@ -42,7 +43,7 @@ PREFIX wikidata: <http://www.wikidata.org/entity/>
 }
 
 :submitterShape {
-    obo:NCIT_C42781 xsd:string * ;
+    obo:NCIT_C42781 xsd:string + ;
     sio:SIO_000116 xsd:string *;
     sio:SIO_000172 xsd:string ?;
     obo:NCIT_C37984 xsd:string ?;
@@ -57,7 +58,7 @@ PREFIX wikidata: <http://www.wikidata.org/entity/>
 :technologyShape {
     obo:OBI_0600047 IRI {0,3} ;
     efo:EFO_0002699 xsd:string ?;
-    obo:FLU_0000848 xsd:float {0,3};
+    obo:FLU_0000848 xsd:double OR xsd:integer {0,3};
     sio:SIO_001167 xsd:string ?;
 }
 
diff --git a/example/maximum_metadata_example.yaml b/example/maximum_metadata_example.yaml
index 0ba9ada..1bc70d7 100644
--- a/example/maximum_metadata_example.yaml
+++ b/example/maximum_metadata_example.yaml
@@ -9,6 +9,7 @@ host:
     host_health_status: http://purl.obolibrary.org/obo/NCIT_C25269
     host_treatment: Process in which the act is intended to modify or alter host status (Compounds)
     host_vaccination: [vaccines1,vaccine2]
+    ethnicity: http://purl.obolibrary.org/obo/HANCESTRO_0010
     additional_host_information: Optional free text field for addtional information
 
 sample:
diff --git a/example/minimal_metadata_example.yaml b/example/minimal_metadata_example.yaml
index 1e8d3f5..51f8a87 100644
--- a/example/minimal_metadata_example.yaml
+++ b/example/minimal_metadata_example.yaml
@@ -5,7 +5,7 @@ host:
 
 sample:
     sample_id: XX
-    collection_date: 2020-01
+    collection_date: "2020-01-01"
     collection_location: http://www.wikidata.org/entity/Q148
 
 virus:
-- 
cgit v1.2.3