From e6d87e15ae7a63147ae59aed8dd61e54b928b84e Mon Sep 17 00:00:00 2001
From: Munyoki Kilyungi
Date: Mon, 11 Sep 2023 14:01:13 +0300
Subject: Add genbank dump.

Signed-off-by: Munyoki Kilyungi <me@bonfacemunyoki.com>
---
 examples/genbank.scm | 80 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 80 insertions(+)
 create mode 100755 examples/genbank.scm

(limited to 'examples')

diff --git a/examples/genbank.scm b/examples/genbank.scm
new file mode 100755
index 0000000..65eed0c
--- /dev/null
+++ b/examples/genbank.scm
@@ -0,0 +1,80 @@
+#! /usr/bin/env guile
+!#
+
+(use-modules (srfi srfi-1)
+             (srfi srfi-26)
+             (rnrs bytevectors)
+             (ice-9 getopt-long)
+             (ice-9 match)
+             (ice-9 regex)
+             (transform strings)
+             (transform sql)
+             (transform triples)
+             (transform special-forms)
+             (transform uuid))
+
+
+
+(define (remap-species-identifiers str)
+  "This procedure remaps identifiers to standard binominal. Obviously this should
+   be sorted by correcting the database!"
+  (match str
+    ["Fly (Drosophila melanogaster dm6)" "Drosophila melanogaster"]
+    ["Oryzias latipes (Japanese medaka)" "Oryzias latipes"]
+    ["Macaca mulatta" "Macaca nemestrina"]
+    ["Bat (Glossophaga soricina)" "Glossophaga soricina"]
+    [str str]))
+
+(define-transformer genbank
+  (tables (Genbank
+           (left-join Species "USING (SpeciesId)")))
+  (schema-triples
+   (gnc:nucleotide a skos:Concept)
+   (gnt:hasSequence rdfs:domain gnc:nucleotide))
+  (triples (ontology
+            'genbank:
+            (field Genbank Id))
+    (set gnt:hasSequence (field Genbank Sequence))
+    (set xkos:classifiedUnder
+         (string->identifier "" (remap-species-identifiers (field Species Fullname))))))
+
+
+
+(let* ((option-spec
+        '((settings (single-char #\s) (value #t))
+          (output (single-char #\o) (value #t))
+          (documentation (single-char #\d) (value #t))))
+       (options (getopt-long (command-line) option-spec))
+       (settings (option-ref options 'settings #f))
+       (output (option-ref options 'output #f))
+       (documentation (option-ref options 'documentation #f))
+       (%connection-settings
+        (call-with-input-file settings
+          read)))
+
+  (with-documentation
+   (name "Genebank Metadata")
+   (connection %connection-settings)
+   (table-metadata? #f)
+   (prefixes
+    '(("rdf:" "<http://www.w3.org/1999/02/22-rdf-syntax-ns#>")
+      ("rdfs:" "<http://www.w3.org/2000/01/rdf-schema#>")
+      ("skos:" "<http://www.w3.org/2004/02/skos/core#>")
+      ("xkos:" "<http://rdf-vocabulary.ddialliance.org/xkos#>")
+      ("gn:" "<http://genenetwork.org/id/>")
+      ("gnc:" "<http://genenetwork.org/category/>")
+      ("gnt:" "<http://genenetwork.org/term/>")
+      ("dct:" "<http://purl.org/dc/terms/>")
+      ("foaf:" "<http://xmlns.com/foaf/0.1/>")
+      ("pubmed:" "<http://rdf.ncbi.nlm.nih.gov/pubmed/>")
+      ("ncbiTaxon:" "<https://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?mode=Info&id=>")
+      ("generif:" "<http://www.ncbi.nlm.nih.gov/gene?cmd=Retrieve&dopt=Graphics&list_uids=>")
+      ("xsd:" "<http://www.w3.org/2001/XMLSchema#>")
+      ("genbank:" "<https://bioregistry.io/reference/genbank:>")
+      ("owl:" "<http://www.w3.org/2002/07/owl#>")))
+   (inputs
+    (list
+     genbank))
+   (outputs
+    `(#:documentation ,documentation
+      #:rdf ,output))))
-- 
cgit v1.2.3