diff options
| -rwxr-xr-x | examples/generif.scm | 159 | ||||
| -rwxr-xr-x | examples/probeset.scm | 10 | ||||
| -rwxr-xr-x | generate-ttl-files.scm | 60 |
3 files changed, 139 insertions, 90 deletions
diff --git a/examples/generif.scm b/examples/generif.scm index fb3208a..894b766 100755 --- a/examples/generif.scm +++ b/examples/generif.scm @@ -16,22 +16,6 @@ -(define (fix-email-id email) - (string-delete #\space email)) - -(define (investigator-attributes->id first-name last-name email) - ;; There is just one record corresponding to "Evan Williams" which - ;; does not have an email ID. To accommodate that record, we - ;; construct the investigator ID from not just the email ID, but - ;; also the first and the last names. It would be preferable to just - ;; find Evan Williams' email ID and insert it into the database. - (string->identifier "investigator" - (string-join - (list first-name last-name (fix-email-id email)) - "_"))) - - - (define-transformer genewiki-symbols (tables (GeneRIF_BASIC) "GROUP BY BINARY symbol") @@ -45,85 +29,87 @@ (set rdfs:label (field GeneRIF_BASIC symbol)))) +;; Some symbols exist in the RIF table that don't exist in the GeneRIF +;; table. +(define-transformer generif-symbols + (tables (GeneRIF) + "WHERE symbol NOT IN (SELECT symbol from GeneRIF_BASIC) GROUP BY BINARY symbol") + (triples + (string->identifier + "symbol" + (regexp-substitute/global #f "[^A-Za-z0-9:]" + (field GeneRIF symbol) + 'pre "_" 'post) + #:proc (lambda (x) x)) + (set rdfs:label + (field GeneRIF symbol)))) + (define-transformer gn-genewiki-entries (tables (GeneRIF (left-join Species "ON Species.SpeciesId = GeneRIF.SpeciesId") (left-join GeneRIFXRef "ON GeneRIFXRef.GeneRIFId = GeneRIF.Id") - (left-join GeneCategory "ON GeneRIFXRef.GeneCategoryId = GeneCategory.Id") - (left-join Investigators "ON Investigators.Email = GeneRIF.email")) - "WHERE GeneRIF.display > 0 AND GeneRIF.VersionId = 0 AND GeneRIF.comment IS NOT NULL GROUP BY GeneRIF.comment, BINARY GeneRIF.symbol") + (left-join GeneCategory "ON GeneRIFXRef.GeneCategoryId = GeneCategory.Id")) + "WHERE GeneRIF.display > 0 AND GeneRIF.comment IS NOT NULL +GROUP BY GeneRIF.Id, GeneRIF.versionId, GeneRIF.symbol, GeneRIF.SpeciesId, +GeneRIF.createtime, GeneRIF.reason") (schema-triples (gnc:GeneWikiEntry a rdfs:Class) (gnc:GNWikiEntry rdfs:subClassOf gnc:GeneWikiEntry) + (gnt:initial a owl:ObjectProperty) + (gnt:initial rdfs:domain gnc:GeneWikiEntry) + (gnt:initial skos:definition "Optional user or project code or your initials") + (gnt:reason a owl:ObjectProperty) + (gnt:reason rdfs:domain gnc:GeneWikiEntry) + (gnt:reason skos:definition "The reason why this resource was modified") (gnc:GNWikiEntry rdfs:comment "Represents GeneRIF Entries entered from GeneNetwork") (gnt:geneSymbol rdfs:domain gnc:GNWikiEntry)) + ;; Here we use the Id and VersionId to uniquely identify comments. + ;; We could use blank-nodes here; however, querying blank nodes + ;; E.g. getting the latest versionId is very complicated. Prefer + ;; normal triplets over blank-nodes. (triples - (string->identifier - "symbol" - (regexp-substitute/global - #f "[^A-Za-z0-9:]" - (field GeneRIF symbol) - 'pre "_" 'post) - #:proc (lambda (x) x)) - (set rdfs:comment - (let* ([generif-comment (sanitize-rdf-string (field GeneRIF comment))] - [create-time (field GeneRIF createtime EntryCreateTime)] - [pmid (field GeneRIF PubMed_ID PMID)] - [web-url (field GeneRIF weburl)] - [species (string->identifier - "" - (remap-species-identifiers (field Species Fullname)) - #:separator "" - #:proc string-capitalize-first)] - [categories - (remove (lambda (x) - (or (eq? x #f) - (and (string? x) - (string-null? x)))) - (remove-duplicates - (string-split-substring - (field ("GROUP_CONCAT(DISTINCT GeneCategory.Name SEPARATOR '$$')" - GeneCategory)) - "$$")))]) - (string->symbol - (string-append - "[ " - (format #f "rdf:type gnc:GNWikiEntry ; ") - (if (string? species) - "" - (format #f "gnt:belongsToSpecies ~a ; " - species)) - (format #f "rdfs:comment ~s^^xsd:string ; " - generif-comment) - (if (string? create-time) - "" - (format #f "dct:created ~s^^xsd:datetime ; " - (time-unix->string - create-time "~5"))) - (if (and (string? pmid) (not (string-null? pmid))) - (format #f - "~{dct:references pubmed:~a ; ~}" - (string-split pmid #\space)) - "") - (if (and (not (string-null? - (string-trim-both (field GeneRIF email)))) - (not (string-null? (field Investigators Email)))) - (format #f "dct:creator ~a ; " - (investigator-attributes->id - (field Investigators FirstName) - (field Investigators LastName) - (field Investigators Email))) - "") - (if (not (null? categories)) - (format #f - "~{gnt:belongsToCategory ~s ; ~}" - categories) - "") - (if (and (string? web-url) (not (string-null? web-url))) - (format #f "foaf:homepage ~s ; " - web-url) - "") - " ] ")))))) + (format #f "gn:wiki-~a-~a" + (field GeneRIF Id) + (field GeneRIF versionId)) + (set rdfs:comment (sanitize-rdf-string (field GeneRIF comment))) + (set rdf:type 'gnc:GNWikiEntry) + (set gnt:symbol + (string->identifier + "symbol" + (regexp-substitute/global #f "[^A-Za-z0-9:]" + (field GeneRIF symbol) + 'pre "_" 'post) + #:proc (lambda (x) x))) + (set dct:created + (let ((create-time (field GeneRIF createtime EntryCreateTime))) + (if (string? create-time) + "" + (annotate-field + (time-unix->string + create-time + "~5") + '^^xsd:datetime)))) + (multiset dct:references + (string-split (field GeneRIF PubMed_ID PMID) + #\space)) + (set foaf:homepage (field GeneRIF weburl)) + (set gnt:belongsToSpecies (string->identifier + "" + (remap-species-identifiers (field Species Fullname)) + #:separator "" + #:proc string-capitalize-first)) + (set dct:hasVersion (annotate-field (format #f "~s" (field GeneRIF versionId)) + '^^xsd:int)) + (set dct:identifier (annotate-field (format #f "~s" (field GeneRIF Id)) + '^^xsd:int)) + (set gnt:initial (sanitize-rdf-string (field GeneRIF initial))) + (set gnt:reason (field GeneRIF reason)) + (set foaf:mbox (sanitize-rdf-string (field GeneRIF email))) + (multiset gnt:belongsToCategory + (string-split + (field ("GROUP_CONCAT(DISTINCT GeneCategory.Name SEPARATOR ';')" + GeneCategory)) + #\;)))) (define-transformer ncbi-genewiki-entries (tables (GeneRIF_BASIC @@ -169,7 +155,7 @@ taxonomic-id)) (format #f "gnt:hasGeneId generif:~a ; " gene-id) - (format #f "gnt:hasVersionId '~a'^^xsd:integer ; " + (format #f "dct:hasVersion '~a'^^xsd:int ; " version-id) (if (and (string? pmid) (not (string-null? pmid))) (format #f @@ -219,6 +205,7 @@ (inputs (list genewiki-symbols + generif-symbols gn-genewiki-entries ncbi-genewiki-entries)) (outputs diff --git a/examples/probeset.scm b/examples/probeset.scm index 9f694af..caf81aa 100755 --- a/examples/probeset.scm +++ b/examples/probeset.scm @@ -79,8 +79,10 @@ (set gnt:hasTargetId (field ("NULLIF(TRIM(ProbeSet.TargetId), '')" TargetId))) - (set gnt:geneSymbol - (field ProbeSet Symbol)) + (multiset gnt:geneSymbol + (map string-trim (string-split + (field ProbeSet Symbol) + #\,))) (set dct:description (sanitize-rdf-string (field ProbeSet description))) (set gnt:targetsRegion (sanitize-rdf-string @@ -119,8 +121,7 @@ "on the minus strand") (else ""))))))) (set gnt:hasGeneId - (ontology 'gene: - (string-trim-both (field ProbeSet GeneId)))) + (field ProbeSet GeneId)) ;; OMIM Link (set dct:references (let ((omim (field ProbeSet OMIM))) @@ -191,6 +192,7 @@ ("rdf:" "<http://www.w3.org/1999/02/22-rdf-syntax-ns#>") ("rdfs:" "<http://www.w3.org/2000/01/rdf-schema#>") ("dct:" "<http://purl.org/dc/terms/>") + ("uniprot:" "<http://purl.uniprot.org/uniprot/>") ("owl:" "<http://www.w3.org/2002/07/owl#>") ("xsd:" "<http://www.w3.org/2001/XMLSchema#>") ("qb:" "<http://purl.org/linked-data/cube#>") diff --git a/generate-ttl-files.scm b/generate-ttl-files.scm new file mode 100755 index 0000000..65db03f --- /dev/null +++ b/generate-ttl-files.scm @@ -0,0 +1,60 @@ +#! ./pre-inst-env +!# +(use-modules (ice-9 format) + (ice-9 futures) + (ice-9 getopt-long) + (ice-9 ftw)) + +(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))) + (define (enter? name stat result) + stat result ;ignore + ;; Skip version control directories if any. + (not (member (basename name) '(".git" ".svn" "CVS")))) + + (define (leaf name stat result) + stat result ;ignore + (when (string-suffix? ".scm" name) + (let* ((base-file-name (basename name ".scm")) + (cmd (format #f " ~a --settings ~a --output ~a --documentation ~a" + name + settings + (string-append output "/" base-file-name ".ttl") + (string-append documentation "/" base-file-name ".md")))) + (touch + (future + (begin + (display (format #f "Running ~a" cmd)) + (display "\n") + (system cmd))))))) + + (define (down name stat result) + name stat ;ignore + result) + + (define (up name stat result) + name stat ;ignore + result) + + (define (skip name stat result) + name stat ;ignore + result) + + ;; Ignore unreadable files/directories but warn the user. + (define (error name stat errno result) + stat ;ignore + (format (current-error-port) "warning: ~a: ~a~%" + name (strerror errno)) + result) + + (file-system-fold enter? leaf down up skip error + 0 ;initial counter is zero bytes + "./examples")) + + |