//add peak to an existing region
region addPeakToRegion(peak nextpeak, region openReg)
{
openReg.name += "," + nextpeak.name;
openReg.score++;
openReg.peakList.Add(nextpeak);
return(openReg);
}
//close an open region
region closeTheRegion(region openReg)
{
openReg.startIndex = openReg.peakList.Min(x => x.startIndex);
openReg.endIndex = openReg.peakList.Max(x => x.endIndex);
openReg.pValue = fishersMethod(openReg.peakList.Select(x => x.pValue).ToList());
//double bonferroniCutoff = cutoffValue / openReg.peakList.Count;
//openReg.peakList.Where(x => x.pValue < bonferroniCutoff).ToList().ForEach(x => x.cutoff = true);
return(openReg);
}
public void updateVariablesLowMemory(region newRegion, List <int> pkDist, List <int> pkregDist, bool narrowThePeak)
{
if (regPerChr.ContainsKey(newRegion.chromosome))
{
regPerChr[newRegion.chromosome]++;
}
else
{
regPerChr.Add(newRegion.chromosome, 1);
}
if (pkPerChr.ContainsKey(newRegion.chromosome))
{
pkPerChr[newRegion.chromosome] += newRegion.peakList.Count;
}
else
{
pkPerChr.Add(newRegion.chromosome, newRegion.peakList.Count);
}
pkInRegDist.AddRange(pkregDist);
allPkDist.AddRange(pkDist);
double tmpValDbl = statistics.addToRegScr(newRegion.score);
regScr[tmpValDbl].numOfPeaks++;
regScr[tmpValDbl].lengths.Add(newRegion.endIndex - newRegion.startIndex);
int tmpVal = statistics.addToRegLen(newRegion.endIndex - newRegion.startIndex);
regLen[tmpVal].numOfPeaks++;
regLen[tmpVal].lengths.Add(newRegion.endIndex - newRegion.startIndex);
regLen[tmpVal].score.Add(newRegion.score);
foreach (peak p in newRegion.peakList)
{
if (_tfsPost.ContainsKey(p.TFname))
{
_tfsPost[p.TFname].numOfPeaks++;
_tfsPost[p.TFname].lengths.Add(p.endIndex - p.startIndex);
}
else
{
_tfsPost.Add(p.TFname, new TF()
{
numOfPeaks = 1,
lengths = new List <int>()
{
p.endIndex - p.startIndex
}
});
}
}
}
public bool filterRegion(region newRegion)
{
if (!filterChromosome(newRegion.chromosome))
{
return(false);
}
if (!filterStart(newRegion.startIndex))
{
return(false);
}
if (!filterEnd(newRegion.endIndex))
{
return(false);
}
if (!filterRegion(newRegion.regionName))
{
return(false);
}
if (!filterTfName(newRegion.peakList.Select(x => x.TFname).ToList()))
{
return(false);
}
if (!filterPeakName(newRegion.peakList.Select(x => x.peakName).ToList()))
{
return(false);
}
if (!filterLowerScore(newRegion.score))
{
return(false);
}
if (!filterHigherScore(newRegion.score))
{
return(false);
}
if (!filterStrand(newRegion.peakList.Select(x => x.strand).ToList(), newRegion.strand))
{
return(false);
}
return(true);
}
public void updateVariables(region newRegion, List <int> pkDist, List <int> pkregDist, bool narrowThePeak, int summitWindow)
{
detectedRegs.Add(newRegion);
pkInRegDist.AddRange(pkregDist);
allPkDist.AddRange(pkDist);
double tmpValDbl = statistics.addToRegScr(newRegion.score);
regScr[tmpValDbl].numOfPeaks++;
regScr[tmpValDbl].lengths.Add(newRegion.endIndex - newRegion.startIndex);
int tmpVal = statistics.addToRegLen(newRegion.endIndex - newRegion.startIndex);
regLen[tmpVal].numOfPeaks++;
regLen[tmpVal].lengths.Add(newRegion.endIndex - newRegion.startIndex);
regLen[tmpVal].score.Add(newRegion.score);
List <Tuple <string, int> > tfsInReg = new List <Tuple <string, int> >();
int cnt = 1;
foreach (peak p in newRegion.peakList)
{
if (cnt < newRegion.peakList.Count)
{
tfsInReg.Add(new Tuple <string, int>(p.TFname, peakStartPlusSummit(p, narrowThePeak, summitWindow)));
prepareNetwork(tfsInReg, newRegion.peakList.ElementAt(cnt));
}
if (_tfsPost.ContainsKey(p.TFname))
{
_tfsPost[p.TFname].numOfPeaks++;
_tfsPost[p.TFname].lengths.Add(p.endIndex - p.startIndex);
}
else
{
_tfsPost.Add(p.TFname, new TF()
{
numOfPeaks = 1,
lengths = new List <int>()
{
p.endIndex - p.startIndex
}
});
}
cnt++;
}
}
//print the region given when the method is called
public void printRegion(region r, StreamWriter outputRegion, StreamWriter outputPeak)
{
printPeak(new peak()
{
chromosome = r.chromosome,
startIndex = r.startIndex,
endIndex = r.endIndex,
name = r.name,
score = r.score,
strand = r.strand,
signalValue = -1,
pValue = (r.pValue == 0.0) ? -1 : r.pValue,
qValue = (r.qValue == 0.0) ? -1 : r.qValue,
summit = -1
}, outputRegion);
if (peakFile)
{
foreach (peak pk in r.peakList)
{
printPeak(pk, outputPeak);
}
}
}
/// <summary>
/// converts an loaded peak to a region. checks if the region name is in the correct format and adds all the peaks in the peak list
/// </summary>
/// <param name="newPeak">the loaded peak (that is actually a region)</param>
/// <param name="lineCounter">the line counter</param>
/// <returns>the loaded region</returns>
public region peakToRegion(peak newPeak, int lineCounter)
{
List <Tuple <string, int> > tfsInReg;
#region convert peak to region
region newRegion = new region()
{
chromosome = newPeak.chromosome,
startIndex = newPeak.startIndex,
endIndex = newPeak.endIndex,
name = newPeak.name,
score = newPeak.score,
strand = newPeak.strand,
signalValue = newPeak.signalValue,
pValue = newPeak.pValue,
qValue = newPeak.qValue,
summit = newPeak.summit,
peakList = new List <peak>()
};
#endregion
#region check if region name correct
if (newRegion.name.Split('-').Length != 2 || !newRegion.name.Split('-').First().StartsWith("reg"))
{
exit("the name field of the input file in line " + lineCounter + " is in a wrong format");
}
else
{
newRegion.regionName = newRegion.name.Split('-').First();
}
#endregion
#region add peaks to peak list
tfsInReg = new List <Tuple <string, int> >();
foreach (string sp in newRegion.name.Split('-').Last().Split(','))
{
if (sp.Split('_').Length < 2)
{
exit("the name field of the input file in line " + lineCounter + " is in a wrong format");
}
else
{
newRegion.peakList.Add(new peak()
{
chromosome = newRegion.chromosome,
startIndex = 1,
endIndex = 1,
TFname = sp.Split('_').First(),
peakName = sp.Split('_').Last(),
name = sp,
score = 0,
strand = '.',
signalValue = 0,
pValue = -1,
qValue = -1,
summit = 1
});
#region add peak statistical data
statistics.addToTfStatsPre(sp.Split('_').First(), new List <int>()
{
0
}, 1);
#endregion
#region keep some stats for network
if (!tfsInReg.Any(x => x.Item1 == sp.Split('_').First()))
{
foreach (Tuple <string, int> s in tfsInReg)
{
if (tfOccs.ContainsKey(s.Item1))
{
if (tfOccs[s.Item1].ContainsKey(sp.Split('_').First()))
{
tfOccs[s.Item1][sp.Split('_').First()].increaseCount(0);
}
else
{
tfOccs[s.Item1].Add(sp.Split('_').First(), new tfOccurrences(0));
}
}
else
{
tfOccs.Add(s.Item1, new Dictionary <string, tfOccurrences>()
{
{ sp.Split('_').First(), new tfOccurrences(0) }
});
}
if (tfOccs.ContainsKey(sp.Split('_').First()))
{
if (tfOccs[sp.Split('_').First()].ContainsKey(s.Item1))
{
tfOccs[sp.Split('_').First()][s.Item1].increaseCount(0);
}
else
{
tfOccs[sp.Split('_').First()].Add(s.Item1, new tfOccurrences(0));
}
}
else
{
tfOccs.Add(sp.Split('_').First(), new Dictionary <string, tfOccurrences>()
{
{ s.Item1, new tfOccurrences(0) }
});
}
}
}
tfsInReg.Add(new Tuple <string, int>(sp.Split('_').First(), 0));
#endregion
}
}
#endregion
return(newRegion);
}
/// <summary>
/// construct a region from an xml file element
/// </summary>
/// <param name="regionElement">xml region element</param>
/// <param name="inputFile">the name of the input xml file</param>
/// <param name="elementCounter">region element counter</param>
/// <param name="lineCounter">the xml file line counter</param>
/// <returns>returns the constructed region</returns>
public region regionFromXML(XElement regionElement, string inputFile, int elementCounter, int lineCounter)
{
int tmpInt; double tmpDbl;
region newRegion = new region();
#region test chromosome 0
try
{
if (!chromosomeNamesAndLength.ContainsKey(regionElement.Attribute("chr").Value) && !ignoreChromosomeLength)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid chromosome entry in element " + elementCounter + " (line " + lineCounter + ")");
}
else
{
newRegion.chromosome = regionElement.Attribute("chr").Value;
}
}
catch (Exception)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " does not contain chromosome information in element " + elementCounter + " (line " + lineCounter + ")");
}
#endregion
#region test start position 1
try
{
if (!int.TryParse(regionElement.Attribute("s").Value, out tmpInt))
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid start index entry in element " + elementCounter + ". Integer expected" + " (line " + lineCounter + ")");
}
if (tmpInt < 0)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid start index entry in element " + elementCounter + ". Positive expected" + " (line " + lineCounter + ")");
}
newRegion.startIndex = tmpInt;
}
catch (Exception)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " does not contain start index information in element " + elementCounter + " (line " + lineCounter + ")");
}
#endregion
#region test end position 2
try
{
if (!int.TryParse(regionElement.Attribute("e").Value, out tmpInt))
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid end index entry in element " + elementCounter + ". Integer expected" + " (line " + lineCounter + ")");
}
if (tmpInt < 0)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid end index entry in element " + elementCounter + ". Positive expected" + " (line " + lineCounter + ")");
}
if (!ignoreChromosomeLength)
{
if (tmpInt > chromosomeNamesAndLength[newRegion.chromosome])
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid end index entry in element " + elementCounter + ". Exceeding chromosome's limits" + " (line " + lineCounter + ")");
}
}
newRegion.endIndex = tmpInt;
}
catch (Exception)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " does not contain end index information in element " + elementCounter + " (line " + lineCounter + ")");
}
#endregion
#region test name 3
try
{
newRegion.name = regionElement.Attribute("n").Value;
if (newRegion.name.Split('-').Length != 2 || !newRegion.name.Split('-').First().StartsWith("reg"))
{
exit("the name field of the input file in line " + lineCounter + " is in a wrong format");
}
else
{
newRegion.regionName = newRegion.name.Split('-').First();
}
}
catch (Exception)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " does not contain name information in element " + elementCounter + " (line " + lineCounter + ")");
}
#endregion
#region test score 4
try
{
if (numOfCols > 4)
{
if (!double.TryParse(regionElement.Attribute("scr").Value, out tmpDbl))
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid score entry in element " + elementCounter + ". Integer expected" + " (line " + lineCounter + ")");
}
newRegion.score = tmpDbl;
}
else
{
newRegion.score = 0;
}
}
catch (Exception)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " does not contain score information in element " + elementCounter + " (line " + lineCounter + ")");
}
#endregion
#region test strand 5
try
{
if (numOfCols > 5)
{
if (!strandSymbols.Exists(x => x == regionElement.Attribute("strd").Value[0]))
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid strand entry in element " + elementCounter + ". +/-/. expected" + " (line " + lineCounter + ")");
}
newRegion.strand = regionElement.Attribute("strd").Value[0];
}
else
{
newRegion.strand = '.';
}
}
catch (Exception)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " does not contain strand information in element " + elementCounter + " (line " + lineCounter + ")");
}
#endregion
#region test pValue 7
try
{
if (numOfCols > 7)
{
if (!double.TryParse(regionElement.Attribute("pv").Value, out tmpDbl))
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid pValue entry in element " + elementCounter + ". Numeric expected" + " (line " + lineCounter + ")");
}
newRegion.pValue = tmpDbl;
}
else
{
newRegion.pValue = -1;
}
}
catch (Exception)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " does not contain pValue information in element " + elementCounter + ". Numeric expected" + " (line " + lineCounter + ")");
}
#endregion
#region test qValue 8
try
{
if (numOfCols > 8)
{
if (!double.TryParse(regionElement.Attribute("qv").Value, out tmpDbl))
{
exit("the file " + inputFile.Split(OSseparator).Last() + " has an invalid qValue entry in element " + elementCounter + ". Numeric expected" + " (line " + lineCounter + ")");
}
newRegion.qValue = tmpDbl;
}
else
{
newRegion.qValue = -1;
}
}
catch (Exception)
{
exit("the file " + inputFile.Split(OSseparator).Last() + " does not contain qValue information in element " + elementCounter + ". Numeric expected" + " (line " + lineCounter + ")");
}
#endregion
return(newRegion);
}
/// <summary>
/// detect and annotate region based on low memory consumption
/// </summary>
/// <returns>data needed for statistics printing</returns>
returnLists regionFinderLowMemory()
{
returnLists rLists = new returnLists();
int pkCounter = 1, realDistance = -1, numOfCols = checkNumberOfFieldsInBedFile(combinedPeakfile);
region newRegion = null;
peak firstPeak, nextPeak;
bool openRegion = false;
FileStream fs = File.Open(@"" + combinedPeakfile, FileMode.Open, FileAccess.Read, FileShare.ReadWrite);
BufferedStream bs = new BufferedStream(fs);
StreamReader sr = new StreamReader(bs);
string firstLine = sr.ReadLine(), nextLine;
Dictionary <string, int> checkIfSorted = new Dictionary <string, int>();
List <int> tmpAllPkDist = new List <int>(), tmpPkInRegDist = new List <int>();
#region printing results
string regionsFileName = resultsDirectory + OSseparator + outfileName + "_regions." + fileType;
StreamWriter outputRegion = new StreamWriter(@"" + regionsFileName);
string peaksInRegionsFileName = resultsDirectory + OSseparator + outfileName + "_regions_peaks." + fileType;
StreamWriter outputPeak = new StreamWriter(@"" + peaksInRegionsFileName);
string xmlFileName = resultsDirectory + OSseparator + outfileName + "_regions.xml";
XmlWriterSettings settings = new XmlWriterSettings();
settings.Indent = true;
settings.IndentChars = "\t";
XmlWriter writer = XmlWriter.Create(@"" + xmlFileName, settings);
writer.WriteStartDocument();
writer.WriteStartElement("regs");
#endregion
#region load the very first peak
if ((firstPeak = peakFromLine(firstLine, numOfCols, combinedPeakfile, pkCounter, null)) == null)
{
firstLine = sr.ReadLine();
pkCounter++;
if ((firstPeak = peakFromLine(firstLine, numOfCols, combinedPeakfile, pkCounter, null)) == null)
{
exit("wrong file format in the combined peak file " + combinedPeakfile.Split(OSseparator).Last());
}
}
//check sorted
if (checkIfSorted.ContainsKey(firstPeak.chromosome))
{
if (checkIfSorted[firstPeak.chromosome] > checkSorting(firstPeak))
{
exit("the combined peak file " + combinedPeakfile.Split(OSseparator).Last() + " is not sorted properly! line: " + pkCounter);
}
else
{
checkIfSorted[firstPeak.chromosome] = checkSorting(firstPeak);
}
}
else
{
checkIfSorted.Add(firstPeak.chromosome, checkSorting(firstPeak));
}
statistics.addToPeaksPerChromosomePre(firstPeak.chromosome, 1);
statistics.addToTfStatsPre(firstPeak.TFname, new List <int>()
{
firstPeak.endIndex - firstPeak.startIndex
}, 1);
#endregion
#region check format
if (numOfCols < 3 || numOfCols > 10)
{
exit("the file " + combinedPeakfile.Split(OSseparator).Last() + " has a non-acceptable format");
}
#endregion
while ((nextLine = sr.ReadLine()) != null)
{
pkCounter++;
nextPeak = peakFromLine(nextLine, numOfCols, combinedPeakfile, pkCounter, null); //as long as you are not at the last element of the peak list you have a nextpeak to assign
#region check for sorting
if (checkIfSorted.ContainsKey(nextPeak.chromosome))
{
if (checkIfSorted[nextPeak.chromosome] > checkSorting(nextPeak))
{
exit("the combined peak file " + combinedPeakfile.Split(OSseparator).Last() + " is not sorted properly! line: " + pkCounter);
}
else
{
checkIfSorted[nextPeak.chromosome] = checkSorting(nextPeak);
}
}
else
{
checkIfSorted.Add(nextPeak.chromosome, checkSorting(nextPeak));
}
statistics.addToPeaksPerChromosomePre(firstPeak.chromosome, 1);
statistics.addToTfStatsPre(firstPeak.TFname, new List <int>()
{
firstPeak.endIndex - firstPeak.startIndex
}, 1);
#endregion
#region change chromosome
if (nextPeak.chromosome != firstPeak.chromosome)
{
if (openRegion) //close the last opened region
{
newRegion = closeTheRegion(newRegion);
if (filterRegion(newRegion))
{
printRegion(newRegion, outputRegion, outputPeak);
writeRegionInXML(writer, newRegion);
rLists.updateVariablesLowMemory(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
openRegion = false;
}
else //you have reached the end of the list and you have no left
{
newRegion = singleRegion(firstPeak, nextRegionCounter());
if (filterRegion(newRegion))
{
printRegion(newRegion, outputRegion, outputPeak);
writeRegionInXML(writer, newRegion);
rLists.updateVariablesLowMemory(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
}
firstPeak = nextPeak;
nextLine = sr.ReadLine();
nextPeak = peakFromLine(nextLine, numOfCols, combinedPeakfile, pkCounter, null);
statistics.addToPeaksPerChromosomePre(firstPeak.chromosome, 1);
statistics.addToTfStatsPre(firstPeak.TFname, new List <int>()
{
firstPeak.endIndex - firstPeak.startIndex
}, 1);
}
#endregion
tmpAllPkDist.Add(realDistance = distanceOfConsecutivePeaks(firstPeak, nextPeak));
switch (strandSpecificRegionDetectionHelp(firstPeak.strand, nextPeak.strand, realDistance, openRegion))
{
case 1:
newRegion = openNewRegion(firstPeak, nextPeak, nextRegionCounter(), '.');
openRegion = true;
tmpPkInRegDist.Add(realDistance);
break;
case 2:
newRegion = addPeakToRegion(nextPeak, newRegion);
tmpPkInRegDist.Add(realDistance);
break;
case 3:
newRegion = singleRegion(firstPeak, nextRegionCounter());
if (filterRegion(newRegion))
{
printRegion(newRegion, outputRegion, outputPeak);
writeRegionInXML(writer, newRegion);
rLists.updateVariablesLowMemory(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
break;
case 4:
newRegion = closeTheRegion(newRegion);
if (filterRegion(newRegion))
{
printRegion(newRegion, outputRegion, outputPeak);
writeRegionInXML(writer, newRegion);
rLists.updateVariablesLowMemory(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
openRegion = false;
break;
default:
exit("something went wrong in region creation");
break;
}
firstPeak = nextPeak;
}
sr.Close();
#region last line of the input file
if (openRegion) //close the last opened region
{
newRegion = closeTheRegion(newRegion);
if (filterRegion(newRegion))
{
printRegion(newRegion, outputRegion, outputPeak);
writeRegionInXML(writer, newRegion);
rLists.updateVariablesLowMemory(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
}
else //you have reached the end of the list and you have no left
{
newRegion = singleRegion(firstPeak, nextRegionCounter());
if (filterRegion(newRegion))
{
printRegion(newRegion, outputRegion, outputPeak);
writeRegionInXML(writer, newRegion);
rLists.updateVariablesLowMemory(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
}
#endregion
rLists.allRegDist.AddRange(rLists.allPkDist.Where(x => x > peakDistance).ToList());
#region closing printings
outputRegion.Close();
outputPeak.Close();
if (!peakFile)
{
File.Delete(@"" + peaksInRegionsFileName);
}
writer.WriteEndElement();
writer.WriteEndDocument();
writer.Close();
if (!xmlFile)
{
File.Delete(@"" + xmlFileName);
}
#endregion
return(rLists);
}
/// <summary>
/// For threads generated for every chromosome
/// </summary>
/// <param name="a"></param>
/// <param name="distance"></param>
/// <param name="allregions"></param>
/// </summary>
/// <returns>data needed for statistics printing</returns>
returnLists regionFinder(List <peak> listOfPeaks)
{
returnLists rLists = new returnLists();
int pkCounter = 1, realDistance = -1, numOfPeaks = listOfPeaks.Count;
region newRegion = null;
peak nextPeak;
bool openRegion = false;
List <int> tmpAllPkDist = new List <int>(), tmpPkInRegDist = new List <int>();
rLists.pkPerChr.Add(listOfPeaks.First().chromosome, 0);
rLists.regPerChr.Add(listOfPeaks.First().chromosome, 0);
foreach (peak examinedPeak in listOfPeaks)
{
#region check for last peak
if (pkCounter != numOfPeaks)
{ //as long as you are not at the last element of the peak list you have a nextpeak to assign
nextPeak = listOfPeaks.ElementAt(pkCounter);
}
else if (openRegion) //close the last opened region
{
newRegion = closeTheRegion(newRegion);
if (filterRegion(newRegion))
{
newRegion.peakList = peakSorting(newRegion.peakList, distanceOption);
rLists.updateVariables(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak, summitWindow);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
break;
}
else //you have reached the end of the list and you have no left
{
newRegion = singleRegion(examinedPeak, nextRegionCounter());
if (filterRegion(newRegion))
{
newRegion.peakList = peakSorting(newRegion.peakList, distanceOption);
rLists.updateVariables(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak, summitWindow);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
break;
}
#endregion
tmpAllPkDist.Add(realDistance = distanceOfConsecutivePeaks(examinedPeak, nextPeak));
switch (strandSpecificRegionDetectionHelp(examinedPeak.strand, nextPeak.strand, realDistance, openRegion))
{
case 1:
newRegion = openNewRegion(examinedPeak, nextPeak, nextRegionCounter(), '.');
openRegion = true;
tmpPkInRegDist.Add(realDistance);
pkCounter++;
break;
case 2:
newRegion = addPeakToRegion(nextPeak, newRegion);
tmpPkInRegDist.Add(realDistance);
pkCounter++;
break;
case 3:
newRegion = singleRegion(examinedPeak, nextRegionCounter());
if (filterRegion(newRegion))
{
newRegion.peakList = peakSorting(newRegion.peakList, distanceOption);
rLists.updateVariables(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak, summitWindow);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
pkCounter++;
break;
case 4:
newRegion = closeTheRegion(newRegion);
if (filterRegion(newRegion))
{
newRegion.peakList = peakSorting(newRegion.peakList, distanceOption);
rLists.updateVariables(newRegion, tmpAllPkDist, tmpPkInRegDist, narrowThePeak, summitWindow);
}
tmpAllPkDist = new List <int>();
tmpPkInRegDist = new List <int>();
openRegion = false;
pkCounter++;
break;
default:
exit("something went wrong in region creation");
break;
}
}
rLists.allRegDist.AddRange(rLists.allPkDist.Where(x => x > peakDistance).ToList());
rLists.pkPerChr[listOfPeaks.First().chromosome] = rLists.detectedRegs.Sum(x => x.peakList.Count);
rLists.regPerChr[listOfPeaks.First().chromosome] = rLists.detectedRegs.Count;
return(rLists);
}
region calculateStatisticData(region examinedRegion, region previousRegion)
{
#region peaksPerChromosome
statistics.addToPeaksPerChromosomePost(examinedRegion.chromosome, examinedRegion.peakList.Count);
#endregion
#region peakInRegionDistance & allPeakDistance & tfStatsPost
int peakIndex = 1;
foreach (peak p in peakSorting(examinedRegion.peakList, true))
{
statistics.allPeaksDistance.Add(1);
if (p.startIndex == 1 && p.endIndex == 1)
{
statistics.peaksInRegionDistance.Add(1);
}
else
{
if (peakIndex < examinedRegion.peakList.Count)
{
statistics.peaksInRegionDistance.Add((examinedRegion.peakList[peakIndex].startIndex + examinedRegion.peakList[peakIndex].summit) - (p.startIndex + p.summit));
}
}
peakIndex++;
statistics.addToTfStatsPost(p.TFname, new List <int>()
{
p.endIndex - p.startIndex
}, 1);
}
#endregion
#region regionsPerChromosome
statistics.addToRegionsPerChromosome(examinedRegion.chromosome, 1);
#endregion
#region regionScore
statistics.addToRegionScore(new List <int>()
{
examinedRegion.endIndex - examinedRegion.startIndex
}, new List <double>()
{
examinedRegion.score
}, 1);
#endregion
#region regionLength
statistics.addToRegionLength(new List <int>()
{
examinedRegion.endIndex - examinedRegion.startIndex
}, new List <double>()
{
examinedRegion.score
}, 1);
#endregion
#region allRegionsDistance
if (previousRegion == null)
{
return(examinedRegion);
}
if (previousRegion.chromosome == examinedRegion.chromosome)
{
statistics.allRegionsDistance.Add(examinedRegion.startIndex - previousRegion.endIndex);
}
#endregion
return(examinedRegion);
}
public void filterRegionsSeriallyXML(string input)
{
string regionsFileName = resultsDirectory + OSseparator + outfileName + "_filtered_regions." + fileType;
StreamWriter outputRegion = new StreamWriter(@"" + regionsFileName);
string peaksInRegionsFileName = resultsDirectory + OSseparator + outfileName + "_filtered_regions_peaks." + fileType;
StreamWriter outputPeak = new StreamWriter(@"" + peaksInRegionsFileName);
string xmlFileName = resultsDirectory + OSseparator + outfileName + "_filtered_regions.xml";
XmlWriterSettings settings = new XmlWriterSettings();
settings.Indent = true;
settings.IndentChars = "\t";
XmlWriter writer = XmlWriter.Create(@"" + xmlFileName, settings);
writer.WriteStartDocument();
writer.WriteStartElement("regs");
XmlReader reader = XmlReader.Create(@"" + inputFile);
XmlReader sTree;
XElement regionElement;
region newRegion, previousRegion = null;
int elementCounter = 1, attributeCounter = 1, lineCounter = 2;
while (reader.ReadToFollowing("reg"))
{
if (reader.IsStartElement())
{
sTree = reader.ReadSubtree();
regionElement = XElement.Load(sTree);
lineCounter++;
newRegion = regionFromXML(regionElement, inputFile, elementCounter, lineCounter);
newRegion.peakList = new List <peak>();
attributeCounter = 1;
foreach (XElement peakElement in regionElement.Descendants("pk"))
{
lineCounter++;
newRegion.peakList.Add(peakFromXML(peakElement, inputFile, elementCounter, attributeCounter, lineCounter));
attributeCounter++;
}
elementCounter++;
lineCounter++;
if (filterRegion(newRegion))
{
previousRegion = calculateStatisticData(newRegion, previousRegion);
printRegion(newRegion, outputRegion, outputPeak);
writeRegionInXML(writer, newRegion);
}
}
else
{
break;
}
}
outputRegion.Close();
outputPeak.Close();
writer.WriteEndElement();
writer.WriteEndDocument();
writer.Close();
if (!peakFile)
{
File.Delete(@"" + peaksInRegionsFileName);
}
if (!xmlFile)
{
File.Delete(@"" + xmlFileName);
}
Console.WriteLine("done!");
}
public void filterRegionSerially(string input)
{
FileStream fs = File.Open(@"" + input, FileMode.Open, FileAccess.Read, FileShare.ReadWrite);
BufferedStream bs = new BufferedStream(fs);
StreamReader sr = new StreamReader(bs);
string line;
int lineCounter = 1;
peak newPeak;
region newRegion, previousRegion = null;
#region check format
int numOfCols = checkNumberOfFieldsInBedFile(input);
if (numOfCols < 3 || numOfCols > 10)
{
exit("the file " + input.Split(OSseparator).Last() + " has a non-acceptable format");
}
#endregion
string regionsFileName = resultsDirectory + OSseparator + outfileName + "_filtered_regions." + fileType;
StreamWriter outputRegion = new StreamWriter(@"" + regionsFileName);
string peaksInRegionsFileName = resultsDirectory + OSseparator + outfileName + "_filtered_regions_peaks." + fileType;
StreamWriter outputPeak = new StreamWriter(@"" + peaksInRegionsFileName);
string xmlFileName = resultsDirectory + OSseparator + outfileName + "_filtered_regions.xml";
XmlWriterSettings settings = new XmlWriterSettings();
settings.Indent = true;
settings.IndentChars = "\t";
XmlWriter writer = XmlWriter.Create(@"" + xmlFileName, settings);
writer.WriteStartDocument();
writer.WriteStartElement("regs");
while ((line = sr.ReadLine()) != null)
{
if ((newPeak = peakFromLine(line, numOfCols, input, lineCounter, null)) == null)
{
continue;
}
newRegion = peakToRegion(newPeak, lineCounter);
if (filterRegion(newRegion))
{
previousRegion = calculateStatisticData(newRegion, previousRegion);
printRegion(newRegion, outputRegion, outputPeak);
writeRegionInXML(writer, newRegion);
}
lineCounter++;
}
outputRegion.Close();
outputPeak.Close();
writer.WriteEndElement();
writer.WriteEndDocument();
writer.Close();
if (!peakFile)
{
File.Delete(@"" + peaksInRegionsFileName);
}
if (!xmlFile)
{
File.Delete(@"" + xmlFileName);
}
Console.WriteLine("done!");
}
/// <summary>
/// Writes a genomic region into the xml file.
/// </summary>
/// <param name="writer">Writer.</param>
/// <param name="reg">Reg.</param>
/// <param name="numOfCols">Number of cols.</param>
public void writeRegionInXML(XmlWriter writer, region reg)
{
#region write region
writer.WriteStartElement("reg");
writer.WriteAttributeString("chr", reg.chromosome);
writer.WriteAttributeString("s", Convert.ToString(reg.startIndex));
writer.WriteAttributeString("e", Convert.ToString(reg.endIndex));
writer.WriteAttributeString("n", reg.name);
writer.WriteAttributeString("scr", Convert.ToString(reg.score));
writer.WriteAttributeString("strd", Convert.ToString(reg.strand));
writer.WriteAttributeString("pv", Convert.ToString(reg.pValue));
writer.WriteAttributeString("qv", Convert.ToString(reg.qValue));
#endregion
#region peak
foreach (peak pk in reg.peakList)
{
writer.WriteStartElement("pk");
switch (numOfCols)
{
case 3:
writer.WriteAttributeString("chr", pk.chromosome);
writer.WriteAttributeString("s", Convert.ToString(pk.startIndex));
writer.WriteAttributeString("e", Convert.ToString(pk.endIndex));
//writer.WriteAttributeString("cut", Convert.ToString(pk.cutoff));
break;
case 4:
writer.WriteAttributeString("chr", pk.chromosome);
writer.WriteAttributeString("s", Convert.ToString(pk.startIndex));
writer.WriteAttributeString("e", Convert.ToString(pk.endIndex));
writer.WriteAttributeString("n", pk.name);
//writer.WriteAttributeString("cut", Convert.ToString(pk.cutoff));
break;
case 5:
writer.WriteAttributeString("chr", pk.chromosome);
writer.WriteAttributeString("s", Convert.ToString(pk.startIndex));
writer.WriteAttributeString("e", Convert.ToString(pk.endIndex));
writer.WriteAttributeString("n", pk.name);
writer.WriteAttributeString("scr", Convert.ToString(pk.score));
//writer.WriteAttributeString("cut", Convert.ToString(pk.cutoff));
break;
case 6:
writer.WriteAttributeString("chr", pk.chromosome);
writer.WriteAttributeString("s", Convert.ToString(pk.startIndex));
writer.WriteAttributeString("e", Convert.ToString(pk.endIndex));
writer.WriteAttributeString("n", pk.name);
writer.WriteAttributeString("scr", Convert.ToString(pk.score));
writer.WriteAttributeString("strd", Convert.ToString(pk.strand));
//writer.WriteAttributeString("cut", Convert.ToString(pk.cutoff));
break;
case 7:
writer.WriteAttributeString("chr", pk.chromosome);
writer.WriteAttributeString("s", Convert.ToString(pk.startIndex));
writer.WriteAttributeString("e", Convert.ToString(pk.endIndex));
writer.WriteAttributeString("n", pk.name);
writer.WriteAttributeString("scr", Convert.ToString(pk.score));
writer.WriteAttributeString("strd", Convert.ToString(pk.strand));
writer.WriteAttributeString("sv", Convert.ToString(pk.signalValue));
//writer.WriteAttributeString("cut", Convert.ToString(pk.cutoff));
break;
case 8:
writer.WriteAttributeString("chr", pk.chromosome);
writer.WriteAttributeString("s", Convert.ToString(pk.startIndex));
writer.WriteAttributeString("e", Convert.ToString(pk.endIndex));
writer.WriteAttributeString("n", pk.name);
writer.WriteAttributeString("scr", Convert.ToString(pk.score));
writer.WriteAttributeString("strd", Convert.ToString(pk.strand));
writer.WriteAttributeString("sv", Convert.ToString(pk.signalValue));
writer.WriteAttributeString("pv", Convert.ToString(pk.pValue));
//writer.WriteAttributeString("cut", Convert.ToString(pk.cutoff));
break;
case 9:
writer.WriteAttributeString("chr", pk.chromosome);
writer.WriteAttributeString("s", Convert.ToString(pk.startIndex));
writer.WriteAttributeString("e", Convert.ToString(pk.endIndex));
writer.WriteAttributeString("n", pk.name);
writer.WriteAttributeString("scr", Convert.ToString(pk.score));
writer.WriteAttributeString("strd", Convert.ToString(pk.strand));
writer.WriteAttributeString("sv", Convert.ToString(pk.signalValue));
writer.WriteAttributeString("pv", Convert.ToString(pk.pValue));
writer.WriteAttributeString("qv", Convert.ToString(pk.qValue));
//writer.WriteAttributeString("cut", Convert.ToString(pk.cutoff));
break;
case 10:
writer.WriteAttributeString("chr", pk.chromosome);
writer.WriteAttributeString("s", Convert.ToString(pk.startIndex));
writer.WriteAttributeString("e", Convert.ToString(pk.endIndex));
writer.WriteAttributeString("n", pk.name);
writer.WriteAttributeString("scr", Convert.ToString(pk.score));
writer.WriteAttributeString("strd", Convert.ToString(pk.strand));
writer.WriteAttributeString("sv", Convert.ToString(pk.signalValue));
writer.WriteAttributeString("pv", Convert.ToString(pk.pValue));
writer.WriteAttributeString("qv", Convert.ToString(pk.qValue));
writer.WriteAttributeString("sm", Convert.ToString(pk.summit));
//writer.WriteAttributeString("cut", Convert.ToString(pk.cutoff));
break;
default:
exit("something went wrong while printing xml");
break;
}
writer.WriteEndElement();
}
#endregion
writer.WriteEndElement();
}
