Skip to content

Commit

Permalink
update variant stat calculation
Browse files Browse the repository at this point in the history
  • Loading branch information
nitin-ebi committed Sep 17, 2024
1 parent bed7e0c commit c4d04a0
Show file tree
Hide file tree
Showing 4 changed files with 126 additions and 10 deletions.
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,7 @@
import org.springframework.context.annotation.Configuration;
import uk.ac.ebi.eva.commons.models.mongo.entity.VariantDocument;
import uk.ac.ebi.eva.pipeline.io.processors.VariantStatsProcessor;
import uk.ac.ebi.eva.pipeline.parameters.InputParameters;

import static uk.ac.ebi.eva.pipeline.configuration.BeanNames.VARIANT_STATS_PROCESSOR;

Expand All @@ -29,7 +30,7 @@ public class VariantStatsProcessorConfiguration {

@Bean(VARIANT_STATS_PROCESSOR)
@StepScope
public ItemProcessor<VariantDocument, VariantDocument> variantStatsProcessor() {
return new VariantStatsProcessor();
public ItemProcessor<VariantDocument, VariantDocument> variantStatsProcessor(InputParameters inputParameters) {
return new VariantStatsProcessor(inputParameters.getStudyId());
}
}
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,6 @@
import java.util.Arrays;
import java.util.Comparator;
import java.util.HashMap;
import java.util.HashSet;
import java.util.List;
import java.util.Map;
import java.util.Optional;
Expand All @@ -30,18 +29,30 @@ public class VariantStatsProcessor implements ItemProcessor<VariantDocument, Var
private static final String DEFAULT_GENOTYPE = "def";
private static final List<String> MISSING_GENOTYPE_ALLELE_REPRESENTATIONS = Arrays.asList(".", "-1");

public VariantStatsProcessor() {
private String studyId;

public VariantStatsProcessor(String studyId) {
this.studyId = studyId;
}

@Override
public VariantDocument process(VariantDocument variant) {
Map<String, Integer> filesIdNumberOfSamplesMap = VariantStatsReader.getFilesIdAndNumberOfSamplesMap();
Set<String> fidSet = filesIdNumberOfSamplesMap.keySet();

String variantRef = variant.getReference();
String variantAlt = variant.getAlternate();
Set<VariantStatsMongo> variantStatsSet = new HashSet<>();

Set<VariantSourceEntryMongo> variantSourceEntrySet = variant.getVariantSources();
// copy the stats that should not be changed/updated and will be copied as it is
Set<VariantStatsMongo> variantStatsSet = variant.getVariantStatsMongo().stream()
.filter(st -> !st.getStudyId().equals(studyId) || !fidSet.contains(st.getFileId()))
.collect(Collectors.toSet());

// get only the ones for which we can calculate the stats
Set<VariantSourceEntryMongo> variantSourceEntrySet = variant.getVariantSources().stream()
.filter(vse -> vse.getStudyId().equals(studyId) && fidSet.contains(vse.getFileId()))
.collect(Collectors.toSet());

for (VariantSourceEntryMongo variantSourceEntry : variantSourceEntrySet) {
String studyId = variantSourceEntry.getStudyId();
String fileId = variantSourceEntry.getFileId();
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -94,10 +94,11 @@ private void populateFilesIdAndNumberOfSamplesMap() {
computed("fid", "$fid"),
computed("numOfSamples", new Document("$size", new Document("$objectToArray", "$samp")))
));
Bson groupStage = group("$fid", sum("totalNumOfSamples", "$numOfSamples"));
Bson groupStage = group("$fid", sum("totalNumOfSamples", "$numOfSamples"), sum("count", 1));
Bson filterStage = match(Filters.eq("count", 1));

filesIdNumberOfSamplesMap = mongoTemplate.getCollection(databaseParameters.getCollectionFilesName())
.aggregate(asList(matchStage, projectStage, groupStage))
.aggregate(asList(matchStage, projectStage, groupStage, filterStage))
.into(new ArrayList<>())
.stream()
.collect(Collectors.toMap(doc -> doc.getString("_id"), doc -> doc.getInteger("totalNumOfSamples")));
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,7 @@
*/
package uk.ac.ebi.eva.pipeline.configuration.jobs.steps;

import com.mongodb.client.MongoCollection;
import org.bson.Document;
import org.junit.After;
import org.junit.Assert;
Expand All @@ -39,6 +40,7 @@
import uk.ac.ebi.eva.utils.EvaJobParameterBuilder;

import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;

import static org.junit.Assert.assertEquals;
Expand Down Expand Up @@ -76,7 +78,6 @@ public class VariantStatsStepTest {
@Before
public void setUp() throws Exception {
mongoRule.getTemporaryDatabase(DATABASE_NAME).drop();
mongoRule.restoreDump(getResourceUrl(MONGO_DUMP), DATABASE_NAME);
}

@After
Expand All @@ -85,7 +86,10 @@ public void cleanUp() {
}

@Test
public void variantStatsStepShouldCalculateAndLoadStats() {
public void variantStatsStepShouldCalculateAndLoadStats() throws Exception {
mongoRule.restoreDump(getResourceUrl(MONGO_DUMP), DATABASE_NAME);


JobParameters jobParameters = new EvaJobParameterBuilder()
.collectionFilesName(COLLECTION_FILES_NAME)
.collectionVariantsName(COLLECTION_VARIANTS_NAME)
Expand Down Expand Up @@ -122,4 +126,103 @@ public void variantStatsStepShouldCalculateAndLoadStats() {
assertEquals(0, variantStats.get("missGt"));
}

@Test
public void variantStatsStepShouldCalculateAndLoadStats_WhereFidHasMoreThanOneFile() {
MongoCollection<Document> filesCollection = mongoRule.getCollection(DATABASE_NAME, COLLECTION_FILES_NAME);
MongoCollection<Document> variantsCollection = mongoRule.getCollection(DATABASE_NAME, COLLECTION_VARIANTS_NAME);

filesCollection.insertMany(Arrays.asList(
new Document("sid", "sid1").append("fid", "fid1").append("fname", "fname11")
.append("samp", new Document("samp11", 0).append("samp12", 1).append("samp13", 2)),
// multiple entries for fid2 in the files collection
new Document("sid", "sid1").append("fid", "fid2").append("fname", "fname21")
.append("samp", new Document("samp21", 0).append("samp22", 1)).append("samp23", 2),
new Document("sid", "sid1").append("fid", "fid2").append("fname", "fname22")
.append("samp", new Document("samp31", 0).append("samp32", 1)).append("samp33", 2)
));

variantsCollection.insertMany(Arrays.asList(
new Document("_id", "chr1_11111111_A_G").append("ref", "A").append("alt", "G")
.append("files", Arrays.asList(
// stats should be calculated
new Document("sid", "sid1").append("fid", "fid1")
.append("samp", new Document("def", "0|0").append("0|1", Arrays.asList(1))),
// should not calculate stats - fid2 has more than one entry in the files collection
new Document("sid", "sid1").append("fid", "fid2")
.append("samp", new Document("def", "0|0").append("0|1", Arrays.asList(1))),
// should not calculate stats - no entry for fid3 in files collection
new Document("sid", "sid1").append("fid", "fid3")
.append("samp", new Document("def", "0|0").append("0|1", Arrays.asList(1))),
// should not calculate stats - different study
new Document("sid", "sid2").append("fid", "fid1")
.append("samp", new Document("def", "0|0").append("0|1", Arrays.asList(1)))
))
.append("st", Arrays.asList(
// should be updated with new values
new Document("sid", "sid1").append("fid", "fid1")
.append("maf", 0.1).append("mgf", 0.1)
.append("mafAl", "A").append("mgfGt", "0|0"),
// should not change as it belongs to different study id
new Document("sid", "sid2").append("fid", "fid1")
.append("maf", 0.20000000298023224).append("mgf", 0.20000000298023224)
.append("mafAl", "A").append("mgfGt", "0|0"),
// should not change as it belongs to different fid
new Document("sid", "sid1").append("fid", "fid3")
.append("maf", 0.30000001192092896).append("mgf", 0.30000001192092896)
.append("mafAl", "A").append("mgfGt", "0|0")

))
));

JobParameters jobParameters = new EvaJobParameterBuilder()
.collectionFilesName(COLLECTION_FILES_NAME)
.collectionVariantsName(COLLECTION_VARIANTS_NAME)
.databaseName(DATABASE_NAME)
.inputStudyId("sid1")
.chunkSize("100")
.toJobParameters();

JobExecution jobExecution = jobLauncherTestUtils.launchStep(BeanNames.VARIANT_STATS_STEP, jobParameters);

// check job completed successfully
assertCompleted(jobExecution);
List<Document> documents = mongoRule.getTemporaryDatabase(DATABASE_NAME).getCollection(COLLECTION_VARIANTS_NAME)
.find().into(new ArrayList<>());
Assert.assertTrue(documents.size() == 1);

// assert data
ArrayList<Document> variantStatsList = documents.stream().filter(doc -> doc.get("_id").equals("chr1_11111111_A_G"))
.findFirst().get().get("st", ArrayList.class);
assertEquals(3, variantStatsList.size());

// assert remained unchanged
Document variantStatsForSid2Fid1 = variantStatsList.stream()
.filter(st -> st.get("sid").equals("sid2") && st.get("fid").equals("fid1")).findFirst().get();
assertEquals(0.20000000298023224, variantStatsForSid2Fid1.get("maf"));
assertEquals(0.20000000298023224, variantStatsForSid2Fid1.get("mgf"));
assertEquals("A", variantStatsForSid2Fid1.get("mafAl"));
assertEquals("0|0", variantStatsForSid2Fid1.get("mgfGt"));

// assert remained unchanged
Document variantStatsForSid1Fid3 = variantStatsList.stream()
.filter(st -> st.get("sid").equals("sid1") && st.get("fid").equals("fid3")).findFirst().get();
assertEquals(0.30000001192092896, variantStatsForSid1Fid3.get("maf"));
assertEquals(0.30000001192092896, variantStatsForSid1Fid3.get("mgf"));
assertEquals("A", variantStatsForSid1Fid3.get("mafAl"));
assertEquals("0|0", variantStatsForSid1Fid3.get("mgfGt"));

// assert updated with new stats
Document variantStatsForSid1Fid1 = variantStatsList.stream()
.filter(st -> st.get("sid").equals("sid1") && st.get("fid").equals("fid1")).findFirst().get();
Document numOfGT = (Document) variantStatsForSid1Fid1.get("numGt");
assertEquals(2, numOfGT.get("0|0"));
assertEquals(1, numOfGT.get("0|1"));
assertEquals(0.1666666716337204, variantStatsForSid1Fid1.get("maf"));
assertEquals(0.3333333432674408, variantStatsForSid1Fid1.get("mgf"));
assertEquals("G", variantStatsForSid1Fid1.get("mafAl"));
assertEquals("0|1", variantStatsForSid1Fid1.get("mgfGt"));
assertEquals(0, variantStatsForSid1Fid1.get("missAl"));
assertEquals(0, variantStatsForSid1Fid1.get("missGt"));
}

}

0 comments on commit c4d04a0

Please sign in to comment.