/**
* Get set up for a run of reporting probesets. Often used to open file
* streams and print headers to files etc.
*
* @param qMethod - Quantification method to be used.
* @param layout - Where the probesets, probes, etc are on the chip.
*
* @return true if success, false otherwise.
*/
bool QuantMethodExprCCCHPReport::prepare(QuantMethod &qMethod, const IntensityMart &iMart)
{
QuantExprMethod *eMethod = dynamic_cast<QuantExprMethod *>(&qMethod);
if (eMethod == NULL) { Err::errAbort("Can only use a QuantMethodExprReport with a QuantExprMethod."); }
setupFileNames(iMart);
int nfiles = m_CHPFileNames.size();
// Make sure our output directory exists.
if (!Fs::isWriteableDir(m_Prefix.c_str()) &&
(Fs::mkdirPath(m_Prefix, false) != APT_OK)) {
APT_ERR_ABORT("Can't make or write to directory: " + m_Prefix);
}
removeAllChps();
// Get CEL file GUIDs
///@todo This be computed by the engine and passed in via AnalysisInfo
m_celGuids.resize(nfiles);
std::string tmp_unc_name;
for (int chip=0; chip<nfiles; chip++) {
FusionCELData cel;
try {
tmp_unc_name=Fs::convertToUncPath(m_CELFileNames[chip]);
cel.SetFileName(tmp_unc_name.c_str());
if (!cel.ReadHeader()) {
Err::errAbort("Unable to read CEL file: "+FS_QUOTE_PATH(tmp_unc_name));
}
GenericData *gdata = cel.GetGenericData();
if (gdata != NULL) {
m_celGuids[chip] = gdata->Header().GetGenericDataHdr()->GetFileId();
}
cel.Close();
}
catch (...) {
Err::errAbort("Unable to read CEL file " + tmp_unc_name);
}
}
int maxProbeSetNameLength = 0;
for (int i=0; i<m_Info.m_ProbesetNames.size(); i++) {
int len = (int)strlen(m_Info.m_ProbesetNames.at(i));
if (m_Info.m_ProbesetDisplayNames.size() > 0 && m_Info.m_ProbesetDisplayNames.at(i) != NULL)
len = (int)strlen(m_Info.m_ProbesetDisplayNames.at(i));
maxProbeSetNameLength = Max(maxProbeSetNameLength, len);
}
// Prepare headers for all CHP files.
wstring algName = StringUtils::ConvertMBSToWCS(m_Info.m_AlgName);
wstring algVersion = StringUtils::ConvertMBSToWCS(m_Info.m_AlgVersion);
wstring chipType = StringUtils::ConvertMBSToWCS(m_Info.m_ChipType);
// For each chip, precreate all probeset signal entries (default to 0.0).
Verbose::out(1,"QuantMethodExprCCCHPReport: Creating temporary files for CHP output");
for (int chip=0; chip<nfiles; chip++) {
try {
ParameterNameValueType param;
// Create tmp chp file
std::string tmp_chp_name=m_CHPFileNames[chip] + ".tmp";
CHPQuantificationData *data = new CHPQuantificationData(tmp_chp_name);
m_TmpChpFiles.push_back(tmp_chp_name);
// set parent header
FusionCELData cel;
try {
tmp_unc_name=Fs::convertToUncPath(m_CELFileNames[chip]);
cel.SetFileName(tmp_unc_name.c_str());
if (!cel.ReadHeader()) {
Err::errAbort("Unable to read CEL file: "+FS_QUOTE_PATH(tmp_unc_name));
}
GenericData *gdata = cel.GetGenericData();
if (gdata != NULL) {
data->GetFileHeader()->GetGenericDataHdr()->AddParent(*gdata->Header().GetGenericDataHdr());
}
cel.Close();
}
catch (...) {
Err::errAbort("Unable to read CEL file: "+FS_QUOTE_PATH(tmp_unc_name));
}
data->SetEntryCount(m_Info.m_NumProbeSets, maxProbeSetNameLength);
data->SetAlgName(algName);
data->SetAlgVersion(algVersion);
data->SetArrayType(chipType);
param.SetName(L"program-name");
param.SetValueText(StringUtils::ConvertMBSToWCS(m_Info.m_ProgramName));
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
param.SetName(L"program-version");
param.SetValueText(StringUtils::ConvertMBSToWCS(m_Info.m_ProgramVersion));
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
param.SetName(L"program-company");
param.SetValueText(StringUtils::ConvertMBSToWCS(m_Info.m_ProgramCompany));
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
// Add algorithm parameters to list.
ParameterNameValueTypeList paramList;
assert(m_Info.m_ParamNames.size() == m_Info.m_ParamValues.size());
for (int i=0; i<m_Info.m_ParamNames.size(); i++) {
if (m_Info.m_ParamValues[i].length() > 0) {
param.SetName(StringUtils::ConvertMBSToWCS(m_Info.m_ParamNames[i]));
param.SetValueText(StringUtils::ConvertMBSToWCS(m_Info.m_ParamValues[i]));
paramList.push_back(param);
}
}
// Add list of all CEL GUIDs in batch
///@todo should this be computed by the engine and passed in via AnalysisInfo?
string prefix = "apt-opt-";
for (int chip=0; chip<m_CHPFileNames.size(); chip++) {
if (m_celGuids[chip].empty() == false) {
string paramName = prefix + "cel-guid-" + ToStr(chip+1);
param.SetName(StringUtils::ConvertMBSToWCS(paramName));
param.SetValueText(StringUtils::ConvertMBSToWCS(m_celGuids[chip]));
paramList.push_back(param);
}
}
data->AddAlgParams(paramList);
// Add the run report parameters to the list
ParameterNameValueTypeList summaryParamList;
std::string blankStr(256, ' ');
for (int source=0; source<m_ChipSummaries.size(); source++) {
ChipSummary::metricDefVec_t metricDefs = m_ChipSummaries[source]->getMetricDefs();
for (int i = 0; i < metricDefs.size(); i++) {
param.SetName(StringUtils::ConvertMBSToWCS(metricDefs[i].m_name));
if (metricDefs[i].m_type == ChipSummary::Metric::Double) {
param.SetValueFloat(-1.0);
}
else if (metricDefs[i].m_type == ChipSummary::Metric::Integer) {
param.SetValueInt32(-1);
}
else if (metricDefs[i].m_type == ChipSummary::Metric::String) {
param.SetValueAscii(blankStr);
}
else {
Err::errAbort("QuantMethodExprCCCHPReport: Unable to handle unknown type: " +
ToStr(metricDefs[i].m_type) );
}
summaryParamList.push_back(param);
}
}
data->AddSummaryParams(summaryParamList);
ProbeSetQuantificationData entry;
CHPQuantificationFileWriter writer(*data);
writer.SeekToDataSet(); // seek to data table location
for (int index=0; index<m_Info.m_ProbesetNames.size(); index++) {
if (m_Info.m_ProbesetDisplayNames.size() > 0 && m_Info.m_ProbesetDisplayNames[index] != NULL)
entry.name = m_Info.m_ProbesetDisplayNames[index];
else
entry.name = m_Info.m_ProbesetNames[index];
entry.quantification = 0.0f;
writer.WriteEntry(entry);
}
delete data;
}
catch (...) {
Err::errAbort("QuantMethodExprCHPReport::prepare() - Unable to write header and/or precreate signal entries to file: " + m_CHPFileNames[chip] + ".tmp");
}
}
// initialize expression signal buffer writer
m_ExpressionQuantificationBufferWriter.Initialize(&m_TmpChpFiles);
return true;
}
void CopyNumberResultWriterTest::testWrite()
{
CopyNumberResultWriter writer;
const char *fileNames[] = {"test1.cn", "test2.cn"};
list<ParameterNameValueType> algParams;
list<ParameterNameValueType> sumParams;
vector<ColumnInfo> cols;
writer.MaximumProbeSetNameLength(12);
writer.MaximumCytoRegionNameLength(12);
writer.MaximumGenotypeProbeSetNameLength(12);
writer.AlgName() = "MYALG";
writer.AlgVersion() = "1.0";
writer.NumberProbeSets() = 10;
writer.NumberCytoRegions() = 10;
writer.NumberGenotypeProbeSets() = 10;
writer.Columns() = cols;
writer.AlgParams() = algParams;
writer.SetChromosomeProbeSetIndexInformation(X_CHR, 0, writer.NumberProbeSets());
for (int i=0; i<2; i++)
{
FusionCELData cel;
cel.SetFileName("../data/small_cel_file");
cel.Read(false);
writer.SummaryParams() = sumParams;
writer.CreateResultFile(cel, fileNames[i]);
cel.Close();
ProbeSetMultiDataCopyNumberData entry;
char buf[64];
for (int j=0; j<writer.NumberProbeSets(); j++)
{
entry.chr = X_CHR;
entry.position = j+i;
snprintf(buf, 64, "%d", j+i);
entry.name = buf;
writer.WriteProbeSetResult(entry);
}
ProbeSetMultiDataCytoRegionData cy;
for (int j=0; j<writer.NumberCytoRegions(); j++)
{
cy.call = 1;
cy.confidenceScore = (float)(j+i);
snprintf(buf, 64, "%d", j+i);
cy.name = buf;
writer.WriteCytoRegionResult(cy);
}
ProbeSetMultiDataGenotypeData gt;
for (int j=0; j<writer.NumberGenotypeProbeSets(); j++)
{
gt.call = 1;
gt.confidence = (float)(j+i);
snprintf(buf, 64, "%d", j+i);
gt.name = buf;
writer.WriteGenotypeProbeSetResult(gt);
}
writer.CloseResultsFile();
}
for (int i=0; i<2; i++)
{
FusionCHPData *chp = FusionCHPDataReg::Read(fileNames[i]);
CPPUNIT_ASSERT(chp != NULL);
FusionCHPMultiDataData *genoChp = FusionCHPMultiDataData::FromBase(chp);
CPPUNIT_ASSERT(genoChp != NULL);
CPPUNIT_ASSERT(genoChp->GetAlgName() == L"MYALG");
CPPUNIT_ASSERT(genoChp->GetAlgVersion() == L"1.0");
CPPUNIT_ASSERT(genoChp->GetArrayType() == L"Hg-small");
CPPUNIT_ASSERT(genoChp->GetEntryCount(ExpressionMultiDataType) == 0);
CPPUNIT_ASSERT(genoChp->GetEntryCount(GenotypeMultiDataType) == 10);
CPPUNIT_ASSERT(genoChp->GetEntryCount(CopyNumberMultiDataType) == 10);
CPPUNIT_ASSERT(genoChp->GetEntryCount(CytoMultiDataType) == 10);
DataSetHeader *dsh = genoChp->GetDataSetHeader(CopyNumberMultiDataType);
CPPUNIT_ASSERT(dsh->GetNameValParamCnt() == 3);
ParameterNameValueTypeConstIt begin;
ParameterNameValueTypeConstIt end;
ParameterNameValueTypeConstIt it;
dsh->GetNameValIterators(begin, end);
it = begin;
CPPUNIT_ASSERT(it->GetValueInt32() == 0);
++it;
CPPUNIT_ASSERT(it->GetValueInt32() == 10);
++it;
CPPUNIT_ASSERT(it->GetValueAscii() == "X");
ProbeSetMultiDataCopyNumberData entry;
char buf[64];
for (int j=0; j<10; j++)
{
genoChp->GetCopyNumberEntry(CopyNumberMultiDataType, j, entry);
CPPUNIT_ASSERT(entry.chr == X_CHR);
CPPUNIT_ASSERT(entry.position == j+i);
snprintf(buf, 64, "%d", j+i);
CPPUNIT_ASSERT(entry.name.compare(buf) == 0);
}
ProbeSetMultiDataCytoRegionData cy;
for (int j=0; j<10; j++)
{
genoChp->GetCytoRegionEntry(CytoMultiDataType, j, cy);
CPPUNIT_ASSERT(cy.call == 1);
CPPUNIT_ASSERT_DOUBLES_EQUAL(cy.confidenceScore, j+i, 0.0001f);
snprintf(buf, 64, "%d", j+i);
CPPUNIT_ASSERT(cy.name.compare(buf) == 0);
}
ProbeSetMultiDataGenotypeData gt;
for (int j=0; j<10; j++)
{
genoChp->GetGenotypeEntry(GenotypeMultiDataType, j, gt);
CPPUNIT_ASSERT(gt.call == 1);
CPPUNIT_ASSERT_DOUBLES_EQUAL(gt.confidence, j+i, 0.0001f);
snprintf(buf, 64, "%d", j+i);
CPPUNIT_ASSERT(gt.name.compare(buf) == 0);
}
delete chp;
}
}
/*! Create a "quantification" CHP file with just the header information. The remainder of the file
* will be created at a later time using the buffer writer technique.
* The CHP file will contain only "quantification" results from an expression analysis.
* @param execId The execution identifier. This identifier is used to identify the batch run that created the CHP files.
* @param celFile The full path to the parent CEL file. The header of the CEL file is copied to the CHP file.
* @param outFile The name of the output CHP file.
* @param probeSetNames The probe set names.
* @param algName The name of the algorithm used to create the results.
* @param algVersion The algorithm version.
* @param chipType the chip type, also known as the probe array type.
* @param programName The name of the program used to create the CHP file.
* @param programVersion The version of the program.
* @param programCompany The company or institution who developed the CHP creating software.
* @param paramNames A list of parameter names to store in the CHP file header.
* @param paramValues A list of parameter values to store in the CHP file header.
* @param sumNames A list of summary statistic names to store in the CHP file header.
* @param sumValues A list of summary statistic values to store in the CHP file header.
*/
static void CreateFileWithHeader
(
const string &execId,
const string &celFile,
const string &outFile,
const list<string> &probeSetNames,
const string &algName,
const string &algVersion,
const string &chipType,
const string &programName,
const string &programVersion,
const string &programCompany,
const vector<string>& paramNames,
const vector<string>& paramValues,
const vector<string>& sumNames,
const vector<string>& sumValues
)
{
// Determine the max probe set name.
int numEntries = (int) probeSetNames.size();
int maxProbeSetNameLength = 0;
for (list<string>::const_iterator it=probeSetNames.begin(); it!=probeSetNames.end(); it++)
{
maxProbeSetNameLength = max(maxProbeSetNameLength, (int) it->length());
}
// Create the data object
CHPQuantificationData *data = new CHPQuantificationData(outFile);
data->SetEntryCount(numEntries, maxProbeSetNameLength);
data->SetAlgName(StringUtils::ConvertMBSToWCS(algName));
data->SetAlgVersion(StringUtils::ConvertMBSToWCS(algVersion));
data->SetArrayType(StringUtils::ConvertMBSToWCS(chipType));
// Store the CEL header
if (celFile.length() > 0 && FileUtils::Exists(celFile.c_str()) == true)
{
FusionCELData cel;
cel.SetFileName(celFile.c_str());
cel.ReadHeader();
GenericData *gdata = cel.GetGenericData();
if (gdata != NULL)
data->GetFileHeader()->GetGenericDataHdr()->AddParent(*gdata->Header().GetGenericDataHdr());
cel.Close();
}
// Add algorithm parameters to list.
ParameterNameValueTypeList params;
ParameterNameValueType param;
if (programName.empty() == false)
{
param.SetName(L"program-name");
param.SetValueText(StringUtils::ConvertMBSToWCS(programName));
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
}
if (programVersion.empty() == false)
{
param.SetName(L"program-version");
param.SetValueText(StringUtils::ConvertMBSToWCS(programVersion));
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
}
if (programCompany.empty() == false)
{
param.SetName(L"program-company");
param.SetValueText(StringUtils::ConvertMBSToWCS(programCompany));
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
}
int nparams = (int) paramNames.size();
param.SetName(L"exec-guid");
param.SetValueAscii(execId);
params.push_back(param);
for(int iparam=0; iparam<nparams; iparam++)
{
param.SetName(StringUtils::ConvertMBSToWCS(paramNames[iparam]));
param.SetValueAscii(paramValues[iparam]);
params.push_back(param);
}
if (params.empty() == false)
data->AddAlgParams(params);
params.clear();
nparams = (int) sumNames.size();
for(int iparam=0; iparam<nparams; iparam++)
{
param.SetName(StringUtils::ConvertMBSToWCS(sumNames[iparam]));
param.SetValueAscii(sumValues[iparam]);
params.push_back(param);
}
if (params.empty() == false)
data->AddSummaryParams(params);
// Creating the writer object will create the file with the header information.
CHPQuantificationFileWriter writer(*data);
// Write the probe set names.
writer.SeekToDataSet();
ProbeSetQuantificationData entry;
for (list<string>::const_iterator it=probeSetNames.begin(); it!=probeSetNames.end(); it++)
{
entry.name = *it;
entry.quantification = 0.0f;
writer.WriteEntry(entry);
}
}
/*! Create a "multi-data" CHP file with just the header information. The remainder of the file
* will be created at a later time using the buffer writer technique.
* The CHP file will contain only "genotyping" results.
* @param execId The execution identifier. This identifier is used to identify the batch run that created the CHP files.
* @param celFile The full path to the parent CEL file. The header of the CEL file is copied to the CHP file.
* @param outFile The name of the output CHP file.
* @param extraColNames The names of the extra data columns. Should not include probe set name, call and confidence columns.
* @param extraColTypes The types (float, int, ubyte) of the extra columns.
* @param numEntries The number of rows (entries) of results to store in the CHP file.
* @param maxProbeSetNameLength The maximum length of the probe set names.
* @param algName The name of the algorithm used to create the results.
* @param algVersion The algorithm version.
* @param chipType the chip type, also known as the probe array type.
* @param programName The name of the program used to create the CHP file.
* @param programVersion The version of the program.
* @param programCompany The company or institution who developed the CHP creating software.
* @param paramNames A list of parameter names to store in the CHP file header.
* @param paramValues A list of parameter values to store in the CHP file header.
* @param sumNames A list of summary statistic names to store in the CHP file header.
* @param sumValues A list of summary statistic values to store in the CHP file header.
*/
static void CreateFileWithHeader
(
const string &execId,
const string &celFile,
const string &outFile,
const vector<string>& extraColNames,
const vector<string>& extraColTypes,
unsigned long numEntries,
int maxProbeSetNameLength,
const string &algName,
const string &algVersion,
const string &chipType,
const string &programName,
const string &programVersion,
const string &programCompany,
const vector<string>& paramNames,
const vector<string>& paramValues,
const vector<string>& sumNames,
const vector<string>& sumValues,
const vector<string>& extraNames,
const vector<string>& extraValues
)
{
// Create the vector of extra columns. The sample code here supports only float, 32 bit integers and 8 bit unsigned integers.
vector<ColumnInfo> extraColumns;
int ncols = (int)extraColNames.size();
for (int icol=0; icol<ncols; icol++)
{
if (extraColTypes[icol] == "float")
{
FloatColumn fcol(StringUtils::ConvertMBSToWCS(extraColNames[icol]));
extraColumns.push_back(fcol);
}
else if (extraColTypes[icol] == "int")
{
IntColumn intcol(StringUtils::ConvertMBSToWCS(extraColNames[icol]));
extraColumns.push_back(intcol);
}
else if (extraColTypes[icol] == "ubyte")
{
UByteColumn ubcol(StringUtils::ConvertMBSToWCS(extraColNames[icol]));
extraColumns.push_back(ubcol);
}
else
{
throw string("Unsupported column type: ") + extraColTypes[icol];
}
}
// Create the data object
CHPMultiDataData *data = new CHPMultiDataData(outFile);
data->SetEntryCount(GenotypeMultiDataType, numEntries, maxProbeSetNameLength, extraColumns);
data->SetAlgName(StringUtils::ConvertMBSToWCS(algName));
data->SetAlgVersion(StringUtils::ConvertMBSToWCS(algVersion));
data->SetArrayType(StringUtils::ConvertMBSToWCS(chipType));
// Store the CEL header
if (celFile.length() > 0 && FileUtils::Exists(celFile.c_str()) == true)
{
FusionCELData cel;
cel.SetFileName(celFile.c_str());
cel.ReadHeader();
GenericData *gdata = cel.GetGenericData();
if (gdata != NULL)
data->GetFileHeader()->GetGenericDataHdr()->AddParent(*gdata->Header().GetGenericDataHdr());
cel.Close();
}
// Add algorithm parameters to list.
ParameterNameValueTypeList params;
ParameterNameValueType param;
if (programName.empty() == false)
{
param.SetName(L"program-name");
param.SetValueText(StringUtils::ConvertMBSToWCS(programName));
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
}
if (programVersion.empty() == false)
{
param.SetName(L"program-version");
param.SetValueText(StringUtils::ConvertMBSToWCS(programVersion));
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
}
if (programCompany.empty() == false)
{
param.SetName(L"program-company");
param.SetValueText(StringUtils::ConvertMBSToWCS(programCompany));
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
}
int nparams = (int) extraNames.size();
for(int iparam=0; iparam<nparams; iparam++)
{
param.SetName(StringUtils::ConvertMBSToWCS(extraNames[iparam]));
param.SetValueAscii(extraValues[iparam]);
data->GetGenericData().Header().GetGenericDataHdr()->AddNameValParam(param);
}
nparams = (int) paramNames.size();
param.SetName(L"exec-guid");
param.SetValueAscii(execId);
params.push_back(param);
for(int iparam=0; iparam<nparams; iparam++)
{
param.SetName(StringUtils::ConvertMBSToWCS(paramNames[iparam]));
param.SetValueAscii(paramValues[iparam]);
params.push_back(param);
}
if (params.empty() == false)
data->AddAlgParams(params);
params.clear();
nparams = (int) sumNames.size();
for(int iparam=0; iparam<nparams; iparam++)
{
param.SetName(StringUtils::ConvertMBSToWCS(sumNames[iparam]));
param.SetValueAscii(sumValues[iparam]);
params.push_back(param);
}
if (params.empty() == false)
data->AddSummaryParams(params);
// Creating the writer object will create the file with the header information.
CHPMultiDataFileWriter writer(*data);
}