From bb1b32ceef948c8d2489eee771411e7e08572621 Mon Sep 17 00:00:00 2001
From: Munyoki Kilyungi
Date: Wed, 19 Jul 2023 20:23:58 +0300
Subject: Dump probesetfreeze metadata using new metadata

---
 examples/dump-probesetfreeze.scm | 77 +++++++++++++++++++---------------------
 1 file changed, 36 insertions(+), 41 deletions(-)

(limited to 'examples')

diff --git a/examples/dump-probesetfreeze.scm b/examples/dump-probesetfreeze.scm
index 0be81ac..a45fd0a 100755
--- a/examples/dump-probesetfreeze.scm
+++ b/examples/dump-probesetfreeze.scm
@@ -16,18 +16,15 @@
   (call-with-input-file (list-ref (command-line) 1)
     read))
 
-(define %dump-directory
-  (list-ref (command-line) 2))
-
 
 (define-dump dump-gene-chip
   (tables (GeneChip))
   (schema-triples
-   (gn:name rdfs:range rdfs:Literal))
+   (gn-term:name rdfs:range rdfs:Literal))
   (triples (string->identifier "platform" (field GeneChip Name))
     (set rdf:type 'gn:platform)
-    (set gn:name (field GeneChip GeneChipName))
-    (set gn:geoPlatform
+    (set gn-term:name (field GeneChip GeneChipName))
+    (set gn-term:geoPlatform
          (ontology 'geoSeries:
                    (string-trim-both (field GeneChip GeoPlatform))))))
 
@@ -41,48 +38,46 @@
            (left-join Tissue "ON ProbeFreeze.TissueId = Tissue.TissueId"))
           "WHERE ProbeSetFreeze.public > 0 AND InfoFiles.InfoPageName IS NULL GROUP BY ProbeFreeze.Id")
   (schema-triples
-   (gn:avgMethod rdfs:range rdfs:Literal)
-   (gn:dataScale rdfs:range rdfs:Literal)
+   (gn-term:avgMethod rdfs:range rdfs:Literal)
+   (gn-term:dataScale rdfs:range rdfs:Literal)
    (gn:probesetDataset rdf:subClassOf gn:dataset))
   (triples
-      (ontology 'probeset:
-                (regexp-substitute/global #f "[^A-Za-z0-9:]"
-                                          (field ProbeSetFreeze Name)
-                                          'pre "_" 'post))
+      (string->identifier
+       ""
+       (regexp-substitute/global
+        #f "[^A-Za-z0-9:]"
+        (field ProbeSetFreeze Name)
+        'pre "_" 'post)
+       #:separator ""
+       #:proc string-capitalize-first)
     (set rdf:type 'gn:probesetDataset)
-    (set gn:avgMethod (string->identifier "avgmethod" (field AvgMethod Name)))
-    (set gn:fullName (field ProbeSetFreeze FullName))
-    (set gn:shortName (field ProbeSetFreeze ShortName))
+    (set gn-term:avgMethod (string->identifier "avgmethod" (field AvgMethod Name)))
+    (set gn-term:fullName (field ProbeSetFreeze FullName))
+    (set gn-term:shortName (field ProbeSetFreeze ShortName))
     (set dct:created (annotate-field
                       (field ProbeSetFreeze CreateTime)
                       '^^xsd:datetime))
-    (set gn:dataScale (field ProbeSetFreeze DataScale))
-    (set gn:tissueName (string->identifier "tissue" (field Tissue Short_Name)))
-    (set gn:datasetOfInbredSet
+    (set gn-term:dataScale (field ProbeSetFreeze DataScale))
+    (set gn-term:tissueName (string->identifier "tissue" (field Tissue Short_Name)))
+    (set gn-term:datasetOfInbredSet
          (string->identifier "inbredSet" (field InbredSet Name InbredSetName)))))
 
 
 
-(call-with-target-database
- %connection-settings
- (lambda (db)
-   (with-output-to-file (string-append %dump-directory "dump-probesetfreeze.ttl")
-     (lambda ()
-       (prefix "dct:" "<http://purl.org/dc/terms/>")
-       (prefix "foaf:" "<http://xmlns.com/foaf/0.1/>")
-       (prefix "generif:" "<http://www.ncbi.nlm.nih.gov/gene?cmd=Retrieve&dopt=Graphics&list_uids=>")
-       (prefix "geoSeries:" "<http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=>")
-       (prefix "gn:" "<http://genenetwork.org/>")
-       (prefix "owl:" "<http://www.w3.org/2002/07/owl#>")
-       (prefix "phenotype:" "<http://genenetwork.org/phenotype/>")
-       (prefix "pubmed:" "<http://rdf.ncbi.nlm.nih.gov/pubmed/>")
-       (prefix "rdf:" "<http://www.w3.org/1999/02/22-rdf-syntax-ns#>")
-       (prefix "rdfs:" "<http://www.w3.org/2000/01/rdf-schema#>")
-       (prefix "uniprot:" "<http://purl.uniprot.org/uniprot/>")
-       (prefix "up:" "<http://purl.uniprot.org/core/>")
-       (prefix "xsd:" "<http://www.w3.org/2001/XMLSchema#>")
-       (prefix "probeset:" "<http://genenetwork.org/probeset/>")
-       (newline)
-       (dump-gene-chip db)
-       (dump-probesetfreeze db))
-     #:encoding "utf8")))
+(dump-with-documentation
+ (name "Probeset freeze metadata")
+ (connection %connection-settings)
+ (table-metadata? #f)
+ (prefixes
+  '(("geoSeries:" "<http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=>")
+    ("gn:" "<http://genenetwork.org/id/>")
+    ("gn-term:" "<http://genenetwork.org/term/>")
+    ("rdf:" "<http://www.w3.org/1999/02/22-rdf-syntax-ns#>")
+    ("rdfs:" "<http://www.w3.org/2000/01/rdf-schema#>")
+    ("xsd:" "<http://www.w3.org/2001/XMLSchema#>")))
+ (inputs
+  (list dump-gene-chip
+        dump-probesetfreeze))
+ (outputs
+  '(#:documentation "./docs/dump-gene-chip.md"
+    #:rdf "./verified-data/dump-probesetfreeze.ttl")))
-- 
cgit v1.2.3