Demo rfc80 poc mutation data count unit tests (cBioPortal#10981)

* Fix CNA table definition

src/main/java/org/cbioportal/web/parameter/GenomicDataFilter.java

@@ -6,6 +6,13 @@ public class GenomicDataFilter extends DataFilter implements Serializable {
     private String hugoGeneSymbol;
     private String profileType;
 
+    public GenomicDataFilter() {}
+
+    public GenomicDataFilter(String hugoGeneSymbol, String profileType) {
+        this.hugoGeneSymbol = hugoGeneSymbol;
+        this.profileType = profileType;
+    }
+
     public String getHugoGeneSymbol() {
         return hugoGeneSymbol;
     }

src/main/resources/db-scripts/clickhouse/clickhouse.sql

@@ -298,18 +298,18 @@ SELECT
     sample_unique_id,
     cancer_study_identifier,
     hugo_gene_symbol,
-    profile_type,
+    replaceOne(stable_id, concat(sd.cancer_study_identifier, '_'), '') as profile_type,
     alteration_value
 FROM
     (SELECT
          sample_id,
          hugo_gene_symbol,
-         profile_type,
+         stable_id,
          alteration_value
     FROM
         (SELECT
             g.hugo_gene_symbol AS hugo_gene_symbol,
-            arrayElement(splitByString('_', assumeNotNull(gp.stable_id)), -1) AS profile_type,
+            gp.stable_id as stable_id,
             arrayMap(x -> (x = '' ? NULL : x), splitByString(',', assumeNotNull(trim(trailing ',' from ga.values)))) AS alteration_value,
             arrayMap(x -> (x = '' ? NULL : toInt32(x)), splitByString(',', assumeNotNull(trim(trailing ',' from gps.ordered_sample_list)))) AS sample_id
         FROM

src/main/resources/org/cbioportal/persistence/mybatisclickhouse/StudyViewFilterMapper.xml

@@ -530,6 +530,7 @@
                         #{studyId}
                     </foreach>
                     AND gpg.gene = #{mutationDataFilter.hugoGeneSymbol}
+                    AND sgp.alteration_type = 'MUTATION_EXTENDED'
             ),
             mutated_samples AS (
                 SELECT DISTINCT sample_unique_id

src/main/resources/org/cbioportal/persistence/mybatisclickhouse/StudyViewMapper.xml

@@ -282,6 +282,7 @@
                 JOIN gene_panel_to_gene_derived gpg ON sgp.gene_panel_id = gpg.gene_panel_id
             WHERE sample_unique_id IN (<include refid="sampleUniqueIdsFromStudyViewFilter"/>)
                 AND gpg.gene = #{genomicDataFilter.hugoGeneSymbol}
+                AND sgp.alteration_type = 'MUTATION_EXTENDED'
         ),
         mutated_count as (
             SELECT count(distinct sample_unique_id)

src/test/java/org/cbioportal/persistence/mybatisclickhouse/GenomicDataFilterTest.java

@@ -0,0 +1,111 @@
+package org.cbioportal.persistence.mybatisclickhouse;
+
+import org.cbioportal.model.GenomicDataCount;
+import org.cbioportal.model.GenomicDataCountItem;
+import org.cbioportal.persistence.helper.StudyViewFilterHelper;
+import org.cbioportal.persistence.mybatisclickhouse.config.MyBatisConfig;
+import org.cbioportal.web.parameter.GenomicDataFilter;
+import org.cbioportal.web.parameter.StudyViewFilter;
+import org.junit.Test;
+import org.junit.runner.RunWith;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.boot.test.autoconfigure.jdbc.AutoConfigureTestDatabase;
+import org.springframework.boot.test.autoconfigure.orm.jpa.DataJpaTest;
+import org.springframework.context.annotation.Import;
+import org.springframework.test.annotation.DirtiesContext;
+import org.springframework.test.context.ContextConfiguration;
+import org.springframework.test.context.junit4.SpringRunner;
+
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+
+import static org.assertj.core.api.Assertions.assertThat;
+
+@RunWith(SpringRunner.class)
+@Import(MyBatisConfig.class)
+@DataJpaTest
+@DirtiesContext
+@AutoConfigureTestDatabase(replace= AutoConfigureTestDatabase.Replace.NONE)
+@ContextConfiguration(initializers = AbstractTestcontainers.Initializer.class)
+public class GenomicDataFilterTest extends AbstractTestcontainers {
+
+    private static final String STUDY_TCGA_PUB = "study_tcga_pub";
+    private static final String STUDY_ACC_TCGA = "acc_tcga";
+
+    @Autowired
+    private StudyViewMapper studyViewMapper;
+
+    @Test
+    public void getCNACounts() {
+        StudyViewFilter studyViewFilter = new StudyViewFilter();
+        studyViewFilter.setStudyIds(List.of(STUDY_TCGA_PUB));
+
+        GenomicDataFilter genomicDataFilterCNA = new GenomicDataFilter("AKT1", "cna");
+        List<GenomicDataCountItem> actualCountsCNA = studyViewMapper.getCNACounts(StudyViewFilterHelper.build(studyViewFilter, null), List.of(genomicDataFilterCNA));
+        List<GenomicDataCountItem> expectedCountsCNA = List.of(
+            new GenomicDataCountItem("AKT1", "cna", List.of(
+                new GenomicDataCount("Homozygously deleted", "-2", 2),
+                new GenomicDataCount("Heterozygously deleted", "-1", 2),
+                new GenomicDataCount("Diploid", "0", 2),
+                new GenomicDataCount("Gained", "1", 2),
+                new GenomicDataCount("Amplified", "2", 2),
+                new GenomicDataCount("NA", "NA", 5)
+            )));
+        assertThat(actualCountsCNA)
+            .usingRecursiveComparison()
+            .ignoringCollectionOrder()
+            .isEqualTo(expectedCountsCNA);
+
+        GenomicDataFilter genomicDataFilterGISTIC = new GenomicDataFilter("AKT1", "gistic");
+        List<GenomicDataCountItem> actualCountsGISTIC = studyViewMapper.getCNACounts(StudyViewFilterHelper.build(studyViewFilter, null), List.of(genomicDataFilterGISTIC));
+        List<GenomicDataCountItem> expectedCountsGISTIC = List.of(
+            new GenomicDataCountItem("AKT1", "gistic", List.of(
+                new GenomicDataCount("Homozygously deleted", "-2", 2),
+                new GenomicDataCount("Heterozygously deleted", "-1", 3),
+                new GenomicDataCount("Diploid", "0", 3),
+                new GenomicDataCount("Gained", "1", 3),
+                new GenomicDataCount("Amplified", "2", 3),
+                new GenomicDataCount("NA", "NA", 1)
+            )));
+        assertThat(actualCountsGISTIC)
+            .usingRecursiveComparison()
+            .ignoringCollectionOrder()
+            .isEqualTo(expectedCountsGISTIC);
+    }
+
+    @Test
+    public void getMutationCounts() {
+        StudyViewFilter studyViewFilter = new StudyViewFilter();
+        studyViewFilter.setStudyIds(List.of(STUDY_TCGA_PUB));
+
+        GenomicDataFilter genomicDataFilterMutation = new GenomicDataFilter("AKT1", "cna");
+        Map<String, Integer> actualMutationCounts = studyViewMapper.getMutationCounts(StudyViewFilterHelper.build(studyViewFilter, null), genomicDataFilterMutation);
+        Map<String, Integer> expectedMutationCounts = new HashMap<>();
+        expectedMutationCounts.put("mutatedCount", 2);
+        expectedMutationCounts.put("notMutatedCount", 2);
+        expectedMutationCounts.put("notProfiledCount", 11);
+        assertThat(actualMutationCounts)
+            .usingRecursiveComparison()
+            .ignoringCollectionOrder()
+            .isEqualTo(expectedMutationCounts);
+    }
+
+    @Test
+    public void getMutationCountsByType() {
+        StudyViewFilter studyViewFilter = new StudyViewFilter();
+        studyViewFilter.setStudyIds(List.of(STUDY_TCGA_PUB));
+
+        GenomicDataFilter genomicDataFilterMutation = new GenomicDataFilter("AKT1", "mutation");
+        List<GenomicDataCountItem> actualMutationCountsByType = studyViewMapper.getMutationCountsByType(StudyViewFilterHelper.build(studyViewFilter, null), List.of(genomicDataFilterMutation));
+        List<GenomicDataCountItem> expectedMutationCountsByType = List.of(
+            new GenomicDataCountItem("AKT1", "mutations", List.of(
+                new GenomicDataCount("nonsense mutation", "nonsense_mutation", 1, 1),
+                new GenomicDataCount("missense mutation", "missense_mutation", 1, 1)
+            )));
+        assertThat(actualMutationCountsByType)
+            .usingRecursiveComparison()
+            .ignoringCollectionOrder()
+            .isEqualTo(expectedMutationCountsByType);
+    }
+}

src/test/java/org/cbioportal/persistence/mybatisclickhouse/StudyViewMapperTest.java

@@ -65,7 +65,7 @@ public void getMutatedGenes() {
             AlterationFilterHelper.build(studyViewFilter.getAlterationFilter()));
         assertEquals(3, alterationCountByGenes.size());
 
-        var testBrca1AlterationCount = alterationCountByGenes.stream().filter(a -> Objects.equals(a.getHugoGeneSymbol(), "brca1")).findFirst();
+        var testBrca1AlterationCount = alterationCountByGenes.stream().filter(a -> Objects.equals(a.getHugoGeneSymbol(), "BRCA1")).findFirst();
         assert (testBrca1AlterationCount.isPresent());
         assertEquals(Integer.valueOf(5), testBrca1AlterationCount.get().getTotalCount());
     }
@@ -118,7 +118,7 @@ public void getTotalProfiledCountsByGene() {
 
         assertEquals(3, totalProfiledCountsMap.size());
 
-        var akt2TotalProfiledCounts = totalProfiledCountsMap.stream().filter(c -> c.getHugoGeneSymbol().equals("akt2")).findFirst();
+        var akt2TotalProfiledCounts = totalProfiledCountsMap.stream().filter(c -> c.getHugoGeneSymbol().equals("AKT2")).findFirst();
         assertTrue(akt2TotalProfiledCounts.isPresent());
         assertEquals(4, akt2TotalProfiledCounts.get().getNumberOfProfiledCases().intValue());
     }