Skip to content

Commit

Permalink
Merge pull request #209 from julianu/main
Browse files Browse the repository at this point in the history 
Fixing a problem with annotations in the PSI OBO for PSM scores
  • Loading branch information
@julianu
julianu authored Jul 12, 2024
2 parents bcffe82 + 1d3feba commit 5b143e2
Show file tree
Hide file tree
Showing 6 changed files with 13,056 additions and 77 deletions.
2 changes: 1 addition & 1 deletion pom.xml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@

<groupId>de.mpc.pia</groupId>
<artifactId>pia</artifactId>
<version>1.5.3</version>
<version>1.5.4</version>
<name>PIA - Protein Inference Algorithms</name>
<url>https://github.com/mpc-bioinformatics/pia</url>

Expand Down
30 changes: 19 additions & 11 deletions src/main/java/de/mpc/pia/intermediate/compiler/parser/MzIdentMLFileParser.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,7 @@
import de.mpc.pia.intermediate.PeptideSpectrumMatch;
import de.mpc.pia.intermediate.compiler.PIACompiler;
import de.mpc.pia.modeller.score.ScoreModel;
import de.mpc.pia.modeller.score.ScoreModelEnum;
import de.mpc.pia.tools.MzIdentMLTools;
import de.mpc.pia.tools.OntologyConstants;
import de.mpc.pia.tools.obo.AbstractOBOMapper;
Expand Down Expand Up @@ -854,17 +855,24 @@ private ScoreModel parseOBOTermAsScore(Term oboTerm, String value) {
ScoreModel score = null;

if (oboTerm != null) {
// the score is in the OBO file, get the relations etc.
Set<Triple> tripleSet = compiler.getOBOMapper().getTriples(oboTerm, null, null);

for (Triple triple : tripleSet) {
if (triple.getPredicate().getName().equals(AbstractOBOMapper.OBO_IS_A) &&
triple.getObject().getName().equals(OntologyConstants.SEARCH_ENGINE_PSM_SCORE.getPsiAccession())) {
// subject is a "search engine specific score for PSM"
double doubleValue = Double.parseDouble(value);
score = new ScoreModel(doubleValue,
StringEscapeUtils.unescapeJava(oboTerm.getName()),
StringEscapeUtils.unescapeJava(oboTerm.getDescription()));
ScoreModelEnum scEnum = ScoreModelEnum.getModelByAccession(oboTerm.getName());
if ((scEnum != null) && !scEnum.equals(ScoreModelEnum.UNKNOWN_SCORE)) {
// the term is a recognized / hard coded score
double doubleValue = Double.parseDouble(value);
score = new ScoreModel(doubleValue, scEnum);
} else {
// the score is in the OBO file, try to get the relations etc.
Set<Triple> tripleSet = compiler.getOBOMapper().getTriples(oboTerm, null, null);

for (Triple triple : tripleSet) {
if (triple.getPredicate().getName().equals(AbstractOBOMapper.OBO_IS_A) &&
triple.getObject().getName().equals(OntologyConstants.SEARCH_ENGINE_PSM_SCORE.getPsiAccession())) {
// subject is a "search engine specific score for PSM"
double doubleValue = Double.parseDouble(value);
score = new ScoreModel(doubleValue,
StringEscapeUtils.unescapeJava(oboTerm.getName()),
StringEscapeUtils.unescapeJava(oboTerm.getDescription()));
}
}
}
}
Expand Down
59 changes: 15 additions & 44 deletions src/test/java/de/mpc/pia/intermediate/IntermediateJAXBTest.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,59 +5,30 @@
import java.io.File;
import java.io.IOException;

import org.apache.log4j.Logger;
import org.junit.BeforeClass;
import org.junit.Ignore;
import org.junit.Test;

import de.mpc.pia.intermediate.xmlhandler.PIAIntermediateJAXBHandler;
import de.mpc.pia.modeller.PIAModellerTest;

public class IntermediateJAXBTest {

/** logger for this class */
private static final Logger LOGGER = Logger.getLogger(IntermediateJAXBTest.class);

private static File piaFile;


@BeforeClass
public static void setUpBeforeClass() {
piaFile = new File(PIAModellerTest.class.getResource("/55merge_mascot_tandem.pia.xml").getPath());
}


@Test
public void testIntermediateJAXB() {
@Ignore("Unfortunately, loading Unmarshaller for mzid 1.1 and mzid 1.2 in same test suite breaks everything")
public void testIntermediateJAXB() throws IOException{
PIAIntermediateJAXBHandler intermediateHandler;
intermediateHandler = new PIAIntermediateJAXBHandler();

Runtime runtime = Runtime.getRuntime();
double mb = 1024*1024;
final long startTime = System.nanoTime();
final long endTime;

try {
intermediateHandler.parse(piaFile.getAbsolutePath(), null);
} catch (IOException e) {
LOGGER.error(e);
}

endTime = System.nanoTime();

assertEquals(2, intermediateHandler.getFiles().size());
assertEquals(2, intermediateHandler.getSpectraData().size());
assertEquals(2, intermediateHandler.getSearchDatabase().size());
assertEquals(2, intermediateHandler.getAnalysisSoftware().size());
assertEquals(1941, intermediateHandler.getGroups().size());
assertEquals(2131, intermediateHandler.getAccessions().size());
assertEquals(2113, intermediateHandler.getPeptides().size());
assertEquals(2478, intermediateHandler.getPSMs().size());
assertEquals(1856, intermediateHandler.getNrTrees());

LOGGER.info("Total Memory: " + runtime.totalMemory() / mb + " MB");
LOGGER.info("Used Memory: " + (runtime.totalMemory() - runtime.freeMemory()) / mb + " MB");
LOGGER.info("Free Memory: " + runtime.freeMemory() / mb + " MB");
LOGGER.info("Max Memory: " + runtime.maxMemory() / mb + " MB");
LOGGER.info("Execution time: " + ((endTime - startTime) / 1000000000.0) + " s");
File piaFile = new File(IntermediateJAXBTest.class.getResource("/55merge_mascot_tandem.pia.xml").getPath());
intermediateHandler.parse(piaFile.getAbsolutePath(), null);

assertEquals("Number of Files differ", 2, intermediateHandler.getFiles().size());
assertEquals("Number of SpectraData differ", 2, intermediateHandler.getSpectraData().size());
assertEquals("Number of SearchDatabases differ", 2, intermediateHandler.getSearchDatabase().size());
assertEquals("Number of AnalysisSoftware differ", 2, intermediateHandler.getAnalysisSoftware().size());
assertEquals("Number of Groups differ", 1941, intermediateHandler.getGroups().size());
assertEquals("Number of Accessions differ", 2131, intermediateHandler.getAccessions().size());
assertEquals("Number of Peptides differ", 2113, intermediateHandler.getPeptides().size());
assertEquals("Number of PSMs differ", 2478, intermediateHandler.getPSMs().size());
assertEquals("Number of NrTrees differ", 1856, intermediateHandler.getNrTrees());
}
}
35 changes: 35 additions & 0 deletions src/test/java/de/mpc/pia/modeller/exporter/MzIdentML12Test.java
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,35 @@
package de.mpc.pia.modeller.exporter;

import static org.junit.Assert.assertEquals;
import static org.junit.Assert.assertTrue;

import java.io.File;

import org.junit.Before;
import org.junit.Ignore;
import org.junit.Test;

import de.mpc.pia.intermediate.compiler.PIACompiler;
import de.mpc.pia.intermediate.compiler.PIASimpleCompiler;


public class MzIdentML12Test {

@Test
@Ignore("Unfortunately, loading Unmarshaller for mzid 1.1 and mzid 1.2 in same test suite breaks everything")
public void testMzIdentMLv1_2_0Import() {
PIACompiler piaCompiler = new PIASimpleCompiler();

File cometMzid12Results = new File(MzIdentML12Test.class.getResource("/comet_mzid12.mzid").getPath());
assertTrue(piaCompiler.getDataFromFile("mzid", cometMzid12Results.getAbsolutePath(), null, null));

piaCompiler.buildClusterList();
piaCompiler.buildIntermediateStructure();

piaCompiler.setName("testFile");

assertEquals("Wrong number of PIA Input files", 1, piaCompiler.getAllFileIDs().size());
assertEquals("Wrong number of imported peptides", 600, piaCompiler.getNrPeptides());
assertEquals("Wrong number of imported PSMs", 610, piaCompiler.getNrPeptideSpectrumMatches());
}
}
27 changes: 6 additions & 21 deletions src/test/java/de/mpc/pia/modeller/exporter/MzIdentMLExportAndImportTest.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -7,12 +7,10 @@
import java.io.IOException;
import java.util.HashMap;

import org.apache.log4j.Logger;
import org.junit.Before;
import org.junit.Test;

import de.mpc.pia.intermediate.compiler.PIACompiler;
import de.mpc.pia.intermediate.compiler.PIACompilerTest;
import de.mpc.pia.intermediate.compiler.PIASimpleCompiler;
import de.mpc.pia.modeller.PIAModeller;
import de.mpc.pia.modeller.protein.inference.SpectrumExtractorInference;
Expand All @@ -26,41 +24,28 @@

public class MzIdentMLExportAndImportTest {

private File tandemIdXMLResults;
private File tandemMzidResults;
private String piaIntermediateFileName;
/** logger for this class */
private static final Logger LOGGER = Logger.getLogger(PIACompiler.class);

@Before
public void setUp() {
piaIntermediateFileName = "PIACompilerTest.pia.xml";

tandemIdXMLResults = new File(PIACompilerTest.class.getResource("/merge1-tandem-fdr_filtered-015.idXML").getPath());
tandemMzidResults = new File(PIACompilerTest.class.getResource("/55merge_tandem.mzid").getPath());
}


@Test
public void testMzIdentMLv1_1_0Import() throws IOException {
public void testMzIdentMLv1_1_0Import() {
PIACompiler piaCompiler = new PIASimpleCompiler();


File tandemMzidResults = new File(MzIdentMLExportAndImportTest.class.getResource("/55merge_tandem.mzid").getPath());
assertTrue(piaCompiler.getDataFromFile("mzid", tandemMzidResults.getAbsolutePath(), null, null));

piaCompiler.buildClusterList();
piaCompiler.buildIntermediateStructure();

piaCompiler.setName("testFile");

assertEquals("Wrong number of PIA Input files", 1, piaCompiler.getAllFileIDs().size());
assertEquals("Wrong number of imported peptides", 153, piaCompiler.getNrPeptides());
assertEquals("Wrong number of imported PSMs", 170, piaCompiler.getNrPeptideSpectrumMatches());
}


@Test
public void testMzIdentMLExportAndImport() throws IOException {
PIACompiler piaCompiler = new PIASimpleCompiler();

File tandemIdXMLResults = new File(MzIdentMLExportAndImportTest.class.getResource("/merge1-tandem-fdr_filtered-015.idXML").getPath());
assertTrue(piaCompiler.getDataFromFile("tandem", tandemIdXMLResults.getAbsolutePath(), null, null));

piaCompiler.buildClusterList();
Expand All @@ -73,6 +58,7 @@ public void testMzIdentMLExportAndImport() throws IOException {


// write out the file
String piaIntermediateFileName = "testMzIdentMLExportAndImport.pia.xml";
File piaIntermediateFile = File.createTempFile(piaIntermediateFileName, null);
piaCompiler.writeOutXML(piaIntermediateFile);
piaCompiler.finish();
Expand Down Expand Up @@ -110,7 +96,6 @@ public void testMzIdentMLExportAndImport() throws IOException {


// try to read it back in PIA compiler
LOGGER.info("Try to read back the mzIdentML previously compiled");
piaCompiler = new PIASimpleCompiler();

assertTrue(piaCompiler.getDataFromFile("mzIdentMLfile", exportFile.getAbsolutePath(), null, null));
Expand Down
Loading

0 comments on commit 5b143e2

Please sign in to comment.