add mock data and config for clickhouse testing on filtered samples a…

…nd mutated genes (cBioPortal#10213) Co-authored-by: Bryan Lai <laib1@mskcc.org>
pughlab · Nov 8, 2023 · c1bdbda · c1bdbda
1 parent 3f7afd4
commit c1bdbda
Show file tree

Hide file tree

Showing 4 changed files with 232 additions and 0 deletions.
diff --git a/...st/java/org/cbioportal/persistence/mybatiscolumnstore/StudyViewMyBatisRepositoryTest.java b/...st/java/org/cbioportal/persistence/mybatiscolumnstore/StudyViewMyBatisRepositoryTest.java
@@ -0,0 +1,49 @@
+package org.cbioportal.persistence.mybatiscolumnstore;
+
+import org.cbioportal.model.AlterationCountByGene;
+import org.cbioportal.model.Sample;
+import org.cbioportal.webparam.StudyViewFilter;
+import org.junit.Assert;
+import org.junit.Test;
+import org.junit.runner.RunWith;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Configurable;
+import org.springframework.test.context.ContextConfiguration;
+import org.springframework.test.context.junit4.SpringJUnit4ClassRunner;
+
+import java.util.ArrayList;
+import java.util.List;
+
+@RunWith(SpringJUnit4ClassRunner.class)
+@ContextConfiguration("/testContextDatabaseClickhouse.xml")
+@Configurable
+public class StudyViewMyBatisRepositoryTest {
+
+    @Autowired
+    private StudyViewMyBatisRepository studyViewMyBatisRepository;
+
+    @Test
+    public void getFilteredSamples() {
+        StudyViewFilter studyViewFilter = new StudyViewFilter();
+
+        List<String> studyIds = new ArrayList<>();
+        studyIds.add("msk_ch_2020");
+        studyViewFilter.setStudyIds(studyIds);
+
+        List<Sample> samples = studyViewMyBatisRepository.getFilteredSamplesFromColumnstore(studyViewFilter);
+        Assert.assertEquals(3, samples.size());
+    }
+
+    @Test
+    public void getMutatedGenes() {
+        StudyViewFilter studyViewFilter = new StudyViewFilter();
+
+        List<String> studyIds = new ArrayList<>();
+        studyIds.add("msk_ch_2020");
+        studyViewFilter.setStudyIds(studyIds);
+
+        List<AlterationCountByGene> mutations = studyViewMyBatisRepository.getMutatedGenes(studyViewFilter);
+        Assert.assertEquals(2, mutations.size());
+    }
+
+}
diff --git a/persistence/persistence-mybatis/src/test/resources/clickhouseSchema.sql b/persistence/persistence-mybatis/src/test/resources/clickhouseSchema.sql
@@ -0,0 +1,121 @@
+create table cna_discrete
+(
+    sample_unique_id          varchar(255),
+    alteration                int,
+    hugo_gene_symbol          varchar(255),
+    gene_panel_stable_id      varchar(255),
+    cancer_study_identifier   varchar(255),
+    genetic_profile_stable_id varchar(255),
+    PRIMARY KEY (sample_unique_id, alteration, hugo_gene_symbol, cancer_study_identifier)
+);
+
+create table genetic_profile
+(
+    sample_unique_id        varchar(255),
+    genetic_alteration_type varchar(255),
+    datatype                varchar(255),
+    value                   varchar(255),
+    cancer_study_identifier varchar(255),
+    PRIMARY KEY (sample_unique_id, genetic_alteration_type, datatype, value, cancer_study_identifier)
+);
+
+create table genetic_profile_counts
+(
+    sample_unique_id          varchar(255),
+    profile_name              varchar(255),
+    genetic_profile_stable_id varchar(255),
+    cancer_study_identifier   varchar(255),
+    count                     int
+);
+
+create table genomic_event
+(
+    sample_unique_id          varchar(255),
+    variant                   varchar(255),
+    hugo_gene_symbol          varchar(255),
+    gene_panel_stable_id      varchar(255),
+    cancer_study_identifier   varchar(255),
+    genetic_profile_stable_id varchar(255),
+    PRIMARY KEY (sample_unique_id, variant, hugo_gene_symbol, cancer_study_identifier, genetic_profile_stable_id)
+);
+
+create table mutation
+(
+    sample_unique_id          varchar(255),
+    variant                   varchar(255),
+    hugo_gene_symbol          varchar(255),
+    gene_panel_stable_id      varchar(255),
+    cancer_study_identifier   varchar(255),
+    genetic_profile_stable_id varchar(255),
+    PRIMARY KEY (sample_unique_id, variant, hugo_gene_symbol, cancer_study_identifier)
+);
+
+create table patient_clinical_attribute_categorical
+(
+    patient_unique_id       varchar(255),
+    attribute_name          varchar(255),
+    attribute_value         varchar(255),
+    cancer_study_identifier varchar(255),
+    PRIMARY KEY (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier)
+);
+
+create table patient_clinical_attribute_numeric
+(
+    patient_unique_id       varchar(255),
+    attribute_name          varchar(255),
+    attribute_value         float,
+    cancer_study_identifier varchar(255),
+    PRIMARY KEY (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier)
+);
+
+create table sample
+(
+    sample_unique_id         varchar(255),
+    sample_stable_id         varchar(255),
+    patient_unique_id        varchar(255),
+    patient_stable_id        varchar(255),
+    cancer_study_identifier  varchar(255),
+    sample_unique_id_base64  varchar(255),
+    patient_unique_id_base64 varchar(255),
+    PRIMARY KEY (sample_unique_id, patient_unique_id, cancer_study_identifier)
+);
+
+create table sample_clinical_attribute_categorical
+(
+    patient_unique_id       varchar(255),
+    sample_unique_id        varchar(255),
+    attribute_name          varchar(255),
+    attribute_value         varchar(255),
+    cancer_study_identifier varchar(255),
+    PRIMARY KEY (patient_unique_id, sample_unique_id, attribute_name, attribute_value, cancer_study_identifier)
+);
+
+create table sample_clinical_attribute_numeric
+(
+    patient_unique_id       varchar(255),
+    sample_unique_id        varchar(255),
+    attribute_name          varchar(255),
+    attribute_value         float,
+    cancer_study_identifier varchar(255),
+    PRIMARY KEY (patient_unique_id, sample_unique_id, attribute_name, attribute_value, cancer_study_identifier)
+);
+
+create table sample_list
+(
+    sample_unique_id        varchar(255),
+    sample_list_stable_id   varchar(255),
+    name                    varchar(255),
+    cancer_study_identifier varchar(255),
+    PRIMARY KEY (sample_unique_id, sample_list_stable_id, name, cancer_study_identifier)
+);
+
+create table structural_variant
+(
+    sample_unique_id          varchar(255),
+    hugo_symbol_gene1         varchar(255),
+    hugo_symbol_gene2         varchar(255),
+    gene_panel_stable_id      varchar(255),
+    cancer_study_identifier   varchar(255),
+    genetic_profile_stable_id varchar(255),
+    PRIMARY KEY (sample_unique_id, hugo_symbol_gene1, hugo_symbol_gene2, cancer_study_identifier)
+);
diff --git a/persistence/persistence-mybatis/src/test/resources/clickhouseTestSql.sql b/persistence/persistence-mybatis/src/test/resources/clickhouseTestSql.sql
@@ -0,0 +1,20 @@
+INSERT INTO patient_clinical_attribute_numeric (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000004', 'AGE', 39.739902, 'msk_ch_2020');
+INSERT INTO patient_clinical_attribute_numeric (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000004', 'TIME_FROM_DX_TO_SEQ', 991, 'msk_ch_2020');
+INSERT INTO patient_clinical_attribute_numeric (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000004', 'TIME_TO_BLOOD_DRAW_FROM_TX', 609, 'msk_ch_2020');
+INSERT INTO patient_clinical_attribute_numeric (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000015', 'AGE', 44.440792, 'msk_ch_2020');
+INSERT INTO patient_clinical_attribute_numeric (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000015', 'TIME_FROM_DX_TO_SEQ', 2558, 'msk_ch_2020');
+INSERT INTO patient_clinical_attribute_numeric (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000015', 'TIME_TO_BLOOD_DRAW_FROM_TX', 5, 'msk_ch_2020');
+INSERT INTO patient_clinical_attribute_numeric (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000023', 'AGE', 61.319645, 'msk_ch_2020');
+INSERT INTO patient_clinical_attribute_numeric (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000023', 'TIME_FROM_DX_TO_SEQ', 245, 'msk_ch_2020');
+INSERT INTO patient_clinical_attribute_numeric (patient_unique_id, attribute_name, attribute_value, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000023', 'TIME_TO_BLOOD_DRAW_FROM_TX', 166, 'msk_ch_2020');
+
+INSERT INTO sample (sample_unique_id, sample_unique_id_base64, sample_stable_id, patient_unique_id, patient_unique_id_base64, patient_stable_id, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000004-N01', '', 'P-0000004-N01', 'msk_ch_2020_P-0000004', '', 'P-0000004', 'msk_ch_2020');
+INSERT INTO sample (sample_unique_id, sample_unique_id_base64, sample_stable_id, patient_unique_id, patient_unique_id_base64, patient_stable_id, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000015-N01', '', 'P-0000015-N01', 'msk_ch_2020_P-0000015', '', 'P-0000015', 'msk_ch_2020');
+INSERT INTO sample (sample_unique_id, sample_unique_id_base64, sample_stable_id, patient_unique_id, patient_unique_id_base64, patient_stable_id, cancer_study_identifier) VALUES ('msk_ch_2020_P-0000023-N01', '', 'P-0000023-N01', 'msk_ch_2020_P-0000023', '', 'P-0000023', 'msk_ch_2020');
+-- 
+INSERT INTO genomic_event (sample_unique_id, variant, hugo_gene_symbol, gene_panel_stable_id, cancer_study_identifier, genetic_profile_stable_id) VALUES ('msk_ch_2020_P-0000004-N01', 'p.R1051Q', 'KDR', '', 'msk_ch_2020', 'msk_ch_2020_mutations');
+INSERT INTO genomic_event (sample_unique_id, variant, hugo_gene_symbol, gene_panel_stable_id, cancer_study_identifier, genetic_profile_stable_id) VALUES ('msk_ch_2020_P-0000004-N01', 'p.T1884I', 'TET2', '', 'msk_ch_2020', 'msk_ch_2020_mutations');
+
+INSERT INTO mutation (sample_unique_id, variant, hugo_gene_symbol, gene_panel_stable_id, cancer_study_identifier, genetic_profile_stable_id) VALUES ('msk_ch_2020_P-0000004-N01', 'p.R1051Q', 'KDR', '', 'msk_ch_2020', 'msk_ch_2020_mutations');
+INSERT INTO mutation (sample_unique_id, variant, hugo_gene_symbol, gene_panel_stable_id, cancer_study_identifier, genetic_profile_stable_id) VALUES ('msk_ch_2020_P-0000004-N01', 'p.T1884I', 'TET2', '', 'msk_ch_2020', 'msk_ch_2020_mutations');
+
diff --git a/persistence/persistence-mybatis/src/test/resources/testContextDatabaseClickhouse.xml b/persistence/persistence-mybatis/src/test/resources/testContextDatabaseClickhouse.xml
@@ -0,0 +1,42 @@
+<?xml version="1.0" encoding="UTF-8"?>
+
+<beans
+    xmlns="http://www.springframework.org/schema/beans"
+    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+    xmlns:util="http://www.springframework.org/schema/util"
+    xmlns:context="http://www.springframework.org/schema/context"
+    xsi:schemaLocation="http://www.springframework.org/schema/beans
+                      http://www.springframework.org/schema/beans/spring-beans.xsd
+                      http://www.springframework.org/schema/context
+                      http://www.springframework.org/schema/context/spring-context.xsd
+                      http://www.springframework.org/schema/util
+                      http://www.springframework.org/schema/util/spring-util.xsd">
+    <bean id="dbcpDataSource" class="org.springframework.jdbc.datasource.DriverManagerDataSource">
+        <property name="url" value="jdbc:h2:mem:;MODE=MySQL;DATABASE_TO_UPPER=false;INIT=RUNSCRIPT FROM 'classpath:clickhouseSchema.sql'\;RUNSCRIPT FROM 'classpath:clickhouseTestSql.sql'"/>
+        <!-- 
+            Enable this to connect to an actual clickhouse Database instead of a mock one.
+            Note that this is only intended for temporary local use, not for automated CI tests.
+        -->         
+<!--        <property name="url" value="jdbc:ch://localhost:8123/cbioportal"/>-->
+<!--        <property name="password" value="cbio_pass"/>-->
+<!--        <property name="username" value="cbio_user"/>-->
+    </bean>
+
+
+    <bean id="sqlSessionFactory" class="org.mybatis.spring.SqlSessionFactoryBean">
+        <property name="dataSource" ref="dbcpDataSource"/>
+        <property name="typeAliasesPackage" value="org.mskcc.cbio.portal.model"/>
+        <property name="typeHandlersPackage" value="org.cbioportal.persistence.mybatis.typehandler"/>
+    </bean>
+
+    <bean id="transactionManager" class="org.springframework.jdbc.datasource.DataSourceTransactionManager">
+        <property name="dataSource" ref="dbcpDataSource" />
+    </bean>
+
+    <bean class="org.mybatis.spring.mapper.MapperScannerConfigurer">
+        <property name="basePackage" value="org.cbioportal.persistence.mybatiscolumnstore"/>
+    </bean>
+
+    <context:component-scan base-package="org.cbioportal.persistence.mybatiscolumnstore"/>
+
+</beans>