@Override
protected void execute() {
try {
// update status
jobStatus.addStatusMessage("10", "Preparing data");
// infrared connector
// DBConnector dbConnector = new DBConnector(species, new File(babelomicsHomePath +
// "/conf/infrared.properties"));
// prepare params
prepare();
// run fatigo's
if (filterList.size() == 0 && !isYourAnnotations) {
throw new ParseException("No biological database selected (eg. --go-bp)");
} else {
// save id lists
IOUtils.write(outdir + "/ranked_list.txt", rankedList.toString());
result.addOutputItem(
new Item(
"ranked_list",
"ranked_list.txt",
"Ranked list",
Item.TYPE.FILE,
Arrays.asList("RANKED_LIST", "CLEAN"),
new HashMap<String, String>(),
"Input data"));
// order ranked list and save genes and rank separately
// FatiScan fatiscan = new
// FatiScan(rankedList,null,dbConnector,FatiScan.DEFAULT_NUMBER_OF_PARTITIONS,testMode,outputFormat,order,duplicatesMode);
FatiScan fatiscan =
new FatiScan(
rankedList,
null,
null,
FatiScan.DEFAULT_NUMBER_OF_PARTITIONS,
testMode,
outputFormat,
order,
duplicatesMode);
fatiscan.prepareLists();
IOUtils.write(outdir + "/id_list.txt", fatiscan.getIdList());
result.addOutputItem(
new Item(
"id_list",
"id_list.txt",
"Id list (sorted)",
Item.TYPE.FILE,
Arrays.asList("IDLIST", "SORTED"),
new HashMap<String, String>(),
"Input data"));
IOUtils.write(outdir + "/statistic.txt", ListUtils.toStringArray(fatiscan.getStatistic()));
result.addOutputItem(
new Item(
"statistic",
"statistic.txt",
"Statistic (sorted)",
Item.TYPE.FILE,
Arrays.asList("STATISTIC", "SORTED"),
new HashMap<String, String>(),
"Input data"));
// significant count
significantCount = new ArrayList<StringBuilder>();
for (int i = 0; i < DEFAULT_PVALUES.length; i++) {
significantCount.add(new StringBuilder("#DB\tNº of significant terms\n"));
}
// annotation report
annotationReport = new StringBuilder();
annotationReport.append("#DB").append("\t").append("Number of annotations").append("\n");
// do fatiscans
double progress = 20;
double inc = 60.0 / filterList.size();
for (FunctionalFilter filter : filterList) {
doTest(rankedList, filter, method);
jobStatus.addStatusMessage("" + progress, "Executing test");
progress += inc;
}
// update status
jobStatus.addStatusMessage("90", "Executing test");
if (isYourAnnotations) {
doYourAnnotationsTest(rankedList, yourAnnotations, method);
}
// update status
jobStatus.addStatusMessage("95", "Saving results");
saveAnnotationsReport();
// save significant count table
for (int i = 0; i < DEFAULT_PVALUES.length; i++) {
IOUtils.write(
outdir + "/significant_count_" + pvalueFormatter.format(DEFAULT_PVALUES[i]) + ".txt",
significantCount.get(i).toString());
}
result
.getOutputItems()
.add(
3,
new Item(
"significant",
"significant_count_${pvalue}.txt",
"Number of significant terms per DB",
Item.TYPE.FILE,
Arrays.asList("TABLE,SUMMARY_TABLE,SIGNIFICANT_COUNT_TABLE"),
new HashMap<String, String>(),
"Significant Results"));
result.addMetaItem(new Item("flags", "SHOW_PVALUES", "", TYPE.MESSAGE));
}
} catch (Exception e) {
e.printStackTrace();
}
}
public void executeAnova() {
DoubleMatrix matrix = null;
List<String> vars = new ArrayList<String>();
List<Integer> indices = new ArrayList<Integer>();
List<String> values = dataset.getVariables().getByName(className).getValues();
if (values.size() == classValues.size()) {
matrix = dataset.getDoubleMatrix();
vars = values;
} else {
for (int i = 0; i < values.size(); i++) {
if (classValues.contains(values.get(i))) {
indices.add(i);
vars.add(values.get(i));
}
}
matrix = dataset.getSubMatrixByColumns(ListUtils.toIntArray(indices));
}
try {
Dataset subDataset = dataset.getSubDataset(className, classValues);
// apply test and multiple test correction according
//
AnovaTest anova = new AnovaTest(matrix, vars);
TestResultList<AnovaTestResult> res = anova.compute();
DiffExpressionUtils.multipleTestCorrection(res, correction);
// create output file
//
int[] columnOrder =
ListUtils.order(subDataset.getVariables().getByName(className).getValues());
int[] rowOrder = ListUtils.order(ArrayUtils.toList(res.getStatistics()), true);
DataFrame dataFrame = new DataFrame(subDataset.getFeatureNames().size(), 0);
dataFrame.addColumn(
"statistic",
ListUtils.toStringList(
ListUtils.ordered(ArrayUtils.toList(res.getStatistics()), rowOrder)));
dataFrame.addColumn(
"p-value",
ListUtils.toStringList(ListUtils.ordered(ArrayUtils.toList(res.getPValues()), rowOrder)));
dataFrame.addColumn(
"adj. p-value",
ListUtils.toStringList(
ListUtils.ordered(ArrayUtils.toList(res.getAdjPValues()), rowOrder)));
dataFrame.setRowNames(ListUtils.ordered(subDataset.getFeatureNames(), rowOrder));
FeatureData featureData = new FeatureData(dataFrame);
File file = new File(outdir + "/" + test + ".txt");
featureData.save(file);
if (file.exists()) {
result.addOutputItem(
new Item(
test + "file",
file.getName(),
"Anova output file",
TYPE.FILE,
new ArrayList<String>(2),
new HashMap<String, String>(2),
"Anova output files"));
}
// getting significative genes
//
DiffExpressionUtils.addSignificativeResults(
subDataset,
test,
"statistic",
res.getStatistics(),
"adj. p-value",
res.getAdjPValues(),
"p-value",
res.getPValues(),
null,
null,
null,
null,
className,
columnOrder,
pValue,
maxDisplay,
this);
DiffExpressionUtils.createFatiScanRedirection(dataFrame, test, "statistic", result, outdir);
} catch (Exception e) {
e.printStackTrace();
abort(
"exception_executeanova_classcomparison",
"error running anova",
"error running anova: " + e.getMessage(),
"error running anova: " + e.getMessage());
}
}
/**
* @throws InvalidColumnIndexException
* @throws IOException
*/
public void executeLimma() {
Limma limma = null;
if (classValues.size() > 2) {
limma = new Limma(babelomicsHomePath + "/bin/diffexp/limma_multiclasses.r");
} else if (classValues.size() == 2) {
limma = new Limma(babelomicsHomePath + "/bin/diffexp/limma_twoclasses.r");
} else if (classValues.size() == 1) {
limma = new Limma(babelomicsHomePath + "/bin/diffexp/limma_oneclass.r");
} else {
abort(
"testmismatched_executelimma_classcomparison",
"test " + test + " not supported for " + classValues.size() + "-class test",
"test " + test + " not supported for " + classValues.size() + "-class test",
"test " + test + " not supported for " + classValues.size() + "-class test");
}
// System.out.println("dataset = " + dataset.toString());
System.out.println("class name = " + className);
limma.setInputFilename(dataset.getDatasetFile().getAbsolutePath());
limma.setClasses(dataset.getVariables().getByName(className).getValues());
limma.setContrast(classValues);
try {
Dataset subDataset = dataset.getSubDataset(className, classValues);
// apply test and multiple test correction according
//
TestResultList<LimmaTestResult> res = limma.compute();
DiffExpressionUtils.multipleTestCorrection(res, correction);
// create output file
//
int[] columnOrder =
ListUtils.order(subDataset.getVariables().getByName(className).getValues());
int[] rowOrder = ListUtils.order(ArrayUtils.toList(res.getStatistics()), true);
DataFrame dataFrame = new DataFrame(subDataset.getFeatureNames().size(), 0);
dataFrame.addColumn(
"statistic",
ListUtils.toStringList(
ListUtils.ordered(ArrayUtils.toList(res.getStatistics()), rowOrder)));
dataFrame.addColumn(
"p-value",
ListUtils.toStringList(ListUtils.ordered(ArrayUtils.toList(res.getPValues()), rowOrder)));
dataFrame.addColumn(
"adj. p-value",
ListUtils.toStringList(
ListUtils.ordered(ArrayUtils.toList(res.getAdjPValues()), rowOrder)));
dataFrame.setRowNames(ListUtils.ordered(subDataset.getFeatureNames(), rowOrder));
FeatureData featureData = new FeatureData(dataFrame);
File file = new File(outdir + "/" + test + ".txt");
featureData.save(file);
if (file.exists()) {
result.addOutputItem(
new Item(
test + "file",
file.getName(),
"Limma output file",
TYPE.FILE,
new ArrayList<String>(2),
new HashMap<String, String>(2),
"Limma output files"));
}
// getting significative genes
//
DiffExpressionUtils.addSignificativeResults(
subDataset,
test,
"statistic",
res.getStatistics(),
"adj. p-value",
res.getAdjPValues(),
"p-value",
res.getPValues(),
null,
null,
null,
null,
className,
columnOrder,
pValue,
maxDisplay,
this);
DiffExpressionUtils.createFatiScanRedirection(dataFrame, test, "statistic", result, outdir);
} catch (Exception e) {
e.printStackTrace();
abort(
"exception_executelimma_classcomparison",
"error running limma",
"error running limma: " + e.toString(),
"error running limma: " + e.toString());
}
}
public void executeT() {
int[] cols = dataset.getColumnIndexesByVariableValue(className, classValues.get(0));
DoubleMatrix sample1 = dataset.getSubMatrixByColumns(cols);
cols = dataset.getColumnIndexesByVariableValue(className, classValues.get(1));
DoubleMatrix sample2 = dataset.getSubMatrixByColumns(cols);
try {
Dataset subDataset = dataset.getSubDataset(className, classValues);
// apply test and multiple test correction according
//
TTest tTest = new TTest();
TestResultList<TTestResult> res = tTest.tTest(sample1, sample2);
DiffExpressionUtils.multipleTestCorrection(res, correction);
// create output file
//
int[] columnOrder =
ListUtils.order(subDataset.getVariables().getByName(className).getValues());
int[] rowOrder = ListUtils.order(ArrayUtils.toList(res.getStatistics()), true);
DataFrame dataFrame = new DataFrame(subDataset.getFeatureNames().size(), 0);
dataFrame.addColumn(
"statistic",
ListUtils.toStringList(
ListUtils.ordered(ArrayUtils.toList(res.getStatistics()), rowOrder)));
dataFrame.addColumn(
"p-value",
ListUtils.toStringList(ListUtils.ordered(ArrayUtils.toList(res.getPValues()), rowOrder)));
dataFrame.addColumn(
"adj. p-value",
ListUtils.toStringList(
ListUtils.ordered(ArrayUtils.toList(res.getAdjPValues()), rowOrder)));
dataFrame.setRowNames(ListUtils.ordered(subDataset.getFeatureNames(), rowOrder));
FeatureData featureData = new FeatureData(dataFrame);
File file = new File(outdir + "/t.txt");
featureData.save(file);
if (file.exists()) {
result.addOutputItem(
new Item(
"tfile",
file.getName(),
"T-test output file",
TYPE.FILE,
new ArrayList<String>(),
new HashMap<String, String>(),
"T-test output files"));
}
// getting significative genes
//
DiffExpressionUtils.addSignificativeResults(
subDataset,
test,
"statistic",
res.getStatistics(),
"adj. p-value",
res.getAdjPValues(),
"p-value",
res.getPValues(),
null,
null,
null,
null,
className,
columnOrder,
pValue,
maxDisplay,
this);
DiffExpressionUtils.createFatiScanRedirection(dataFrame, test, "statistic", result, outdir);
} catch (Exception e) {
e.printStackTrace();
abort("exception_executet_classcomparison", "ERROR", "Error running t-test", "");
}
}