update to new data file. #532

opencb · Mar 16, 2020 · 175c922 · 175c922
1 parent 7da9273
commit 175c922
Show file tree

Hide file tree

Showing 3 changed files with 8 additions and 4 deletions.
diff --git a/cellbase-core/src/main/resources/configuration.yml b/cellbase-core/src/main/resources/configuration.yml
@@ -101,7 +101,7 @@ download:
  gwasCatalog:
   host: ftp://ftp.ebi.ac.uk/pub/databases/gwas/releases/2016/09/28/gwas-catalog-associations.tsv
  hpo:
-  host: http://compbio.charite.de/hudson/job/hpo.annotations.monthly/lastStableBuild/artifact/annotation/ALL_SOURCES_ALL_FREQUENCIES_diseases_to_genes_to_phenotypes.txt
+  host: http://compbio.charite.de/hudson/job/hpo.annotations.monthly/lastStableBuild/artifact/annotation/phenotype_to_genes.txt
  disgenet:
   host: https://www.disgenet.org/static/disgenet_ap1/files/downloads
   files:

diff --git a/cellbase-lib/src/main/java/org/opencb/cellbase/lib/builders/GeneParser.java b/cellbase-lib/src/main/java/org/opencb/cellbase/lib/builders/GeneParser.java
@@ -98,7 +98,7 @@ public GeneParser(Path geneDirectoryPath, Path genomeSequenceFastaFile,
                 geneDirectoryPath.resolve("idmapping_selected.tab.gz"), geneDirectoryPath.resolve("motif_features.gff.gz"),
                 geneDirectoryPath.getParent().getParent().resolve("common/expression/allgenes_updown_in_organism_part.tab.gz"),
                 geneDirectoryPath.resolve("dgidb.tsv"),
-                geneDirectoryPath.resolve("ALL_SOURCES_ALL_FREQUENCIES_diseases_to_genes_to_phenotypes.txt"),
+                geneDirectoryPath.resolve("phenotype_to_genes.txt"),
                 geneDirectoryPath.resolve("all_gene_disease_associations.txt.gz"),
                 geneDirectoryPath.resolve("gnomad.v2.1.1.lof_metrics.by_transcript.txt.bgz"),
                 genomeSequenceFastaFile, speciesConfiguration, flexibleGTFParsing, serializer);

diff --git a/cellbase-lib/src/main/java/org/opencb/cellbase/lib/builders/GeneParserUtils.java b/cellbase-lib/src/main/java/org/opencb/cellbase/lib/builders/GeneParserUtils.java
@@ -214,9 +214,13 @@ public static Map<String, List<GeneTraitAssociation>> getGeneDiseaseAssociationM
             bufferedReader.readLine();
             while ((line = bufferedReader.readLine()) != null) {
                 fields = line.split("\t");
+                String omimId = fields[6];
+                String geneSymbol = fields[3];
+                String hpoId = fields[0];
+                String diseaseName = fields[1];
                 GeneTraitAssociation disease =
-                        new GeneTraitAssociation(fields[0], fields[4], fields[3], 0f, 0, new ArrayList<>(), new ArrayList<>(), "hpo");
-                addValueToMapElement(geneDiseaseAssociationMap, fields[1], disease);
+                        new GeneTraitAssociation(omimId, diseaseName, hpoId, 0f, 0, new ArrayList<>(), new ArrayList<>(), "hpo");
+                addValueToMapElement(geneDiseaseAssociationMap, geneSymbol, disease);
             }
             bufferedReader.close();
         }