Add entrez_id to alteration Count endpoints (#10936)

cBioPortal · Aug 13, 2024 · d7af5e5 · d7af5e5
1 parent 280dd58
commit d7af5e5
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 7 deletions.
diff --git a/src/main/resources/db-scripts/clickhouse/clickhouse.sql b/src/main/resources/db-scripts/clickhouse/clickhouse.sql
@@ -81,6 +81,7 @@ CREATE TABLE IF NOT EXISTS genomic_event_derived
 (
     sample_unique_id          String,
     hugo_gene_symbol          String,
+    entrez_gene_id            Int32,
     gene_panel_stable_id      LowCardinality(String),
     cancer_study_identifier   LowCardinality(String),
     genetic_profile_stable_id LowCardinality(String),
@@ -95,12 +96,13 @@ CREATE TABLE IF NOT EXISTS genomic_event_derived
     sv_event_info             String,
     patient_unique_id         String
 ) ENGINE = MergeTree
-      ORDER BY ( variant_type, hugo_gene_symbol, genetic_profile_stable_id, sample_unique_id);
+      ORDER BY ( variant_type, entrez_gene_id, hugo_gene_symbol, genetic_profile_stable_id, sample_unique_id);
 
 INSERT INTO genomic_event_derived
 -- Insert Mutations
 SELECT concat(cs.cancer_study_identifier, '_', sample.stable_id) AS sample_unique_id,
        gene.hugo_gene_symbol                                     AS hugo_gene_symbol,
+       gene.entrez_gene_id                                       AS entrez_gene_id,
        ifNull(gp.stable_id, 'WES')                               AS gene_panel_stable_id,
        cs.cancer_study_identifier                                AS cancer_study_identifier,
        g.stable_id                                               AS genetic_profile_stable_id,
@@ -128,6 +130,7 @@ UNION ALL
 -- Insert CNA Genes
 SELECT concat(cs.cancer_study_identifier, '_', sample.stable_id) AS sample_unique_id,
        gene.hugo_gene_symbol                                     AS hugo_gene_symbol,
+       gene.entrez_gene_id                                       AS entrez_gene_id,
        ifNull(gp.stable_id, 'WES')                               AS gene_panel_stable_id,
        cs.cancer_study_identifier                                AS cancer_study_identifier,
        g.stable_id                                               AS genetic_profile_stable_id,
@@ -155,6 +158,7 @@ UNION ALL
 -- Insert Structural Variants Site1
 SELECT concat(cs.cancer_study_identifier, '_', s.stable_id) AS sample_unique_id,
        gene.hugo_gene_symbol                                AS hugo_gene_symbol,
+       gene.entrez_gene_id                                  AS entrez_gene_id,
        ifNull(gene_panel.stable_id, 'WES')                  AS gene_panel_stable_id,
        cs.cancer_study_identifier                           AS cancer_study_identifier,
        gp.stable_id                                         AS genetic_profile_stable_id,
@@ -180,6 +184,7 @@ UNION ALL
 -- Insert Structural Variants Site2
 SELECT concat(cs.cancer_study_identifier, '_', s.stable_id) AS sample_unique_id,
        gene.hugo_gene_symbol                                AS hugo_gene_symbol,
+       gene.entrez_gene_id                                  AS entrez_gene_id,
        ifNull(gene_panel.stable_id, 'WES')                  AS gene_panel_stable_id,
        cs.cancer_study_identifier                           AS cancer_study_identifier,
        gp.stable_id                                         AS genetic_profile_stable_id,

diff --git a/src/main/resources/org/cbioportal/persistence/mybatisclickhouse/StudyViewMapper.xml b/src/main/resources/org/cbioportal/persistence/mybatisclickhouse/StudyViewMapper.xml
@@ -24,7 +24,7 @@
     <select id="getMutatedGenes" resultType="org.cbioportal.model.AlterationCountByGene">
         SELECT
             hugo_gene_symbol as hugoGeneSymbol,
-            1 as entrezGeneId,
+            entrez_gene_id as entrezGeneId,
             COUNT(DISTINCT sample_unique_id) as numberOfAlteredCases,
             COUNT(*) as totalCount
         FROM genomic_event_derived
@@ -37,14 +37,14 @@
                 <include refid="mutationAlterationFilter"/>
             </if>
         </where>
-        GROUP BY hugo_gene_symbol;
+        GROUP BY entrez_gene_id, hugo_gene_symbol;
     </select>
 
     <!-- /cna-genes/fetch (returns CopyNumberCountByGene) -->
     <select id="getCnaGenes" resultType="org.cbioportal.model.CopyNumberCountByGene">
         SELECT
         hugo_gene_symbol as hugoGeneSymbol,
-        1 as entrezGeneId,
+        entrez_gene_id as entrezGeneId,
         cna_alteration as alteration,
         cna_cytoband as cytoband,
         COUNT(DISTINCT sample_unique_id) as numberOfAlteredCases,
@@ -59,13 +59,13 @@
                 <include refid="cnaAlterationFilter"/>
             </if>
         </where>
-        GROUP BY hugo_gene_symbol, alteration, cytoband;
+        GROUP BY entrez_gene_id, hugo_gene_symbol, alteration, cytoband;
     </select>
 
     <select id="getStructuralVariantGenes" resultType="org.cbioportal.model.AlterationCountByGene">
         SELECT
         hugo_gene_symbol as hugoGeneSymbol,
-        1 as entrezGeneId,
+        entrez_gene_id as entrezGeneId,
         COUNT(DISTINCT sample_unique_id) as numberOfAlteredCases,
         COUNT(*) as totalCount
         FROM genomic_event_derived
@@ -75,7 +75,7 @@
                 <property name="filter_type" value="'SAMPLE_AND_PATIENT_ID'"/>
             </include>
         </where>
-        GROUP BY hugo_gene_symbol;
+        GROUP BY entrez_gene_id, hugo_gene_symbol;
     </select>
 
     <select id="getSampleClinicalDataFromStudyViewFilter" resultType="org.cbioportal.model.ClinicalData">