public static List <double[]> HeatMapPearsonCorrelation(string readfilePath, string outpath)
{
List <string[]> meanList = FileProcess.ReadCSV(readfilePath);
List <double[]> pclist = new List <double[]>();
for (int i = 0; i < meanList.Count; i++)
{
double[] pcline = new double[meanList.Count];
double[] line = new double[meanList[i].GetLength(0)];
for (int m = 0; m < meanList[i].GetLength(0); m++)
{
line[m] = double.Parse(meanList[i][m]);
}
for (int j = 0; j < meanList.Count; j++)
{
double[] lineX = new double[meanList[j].GetLength(0)];
for (int n = 0; n < meanList[j].GetLength(0); n++)
{
lineX[n] = double.Parse(meanList[j][n]);
}
double pc = Cal_Pearson_Correlation.CalPearsonCorrelation(line, lineX);
pcline[j] = pc;
}
pclist.Add(pcline);
}
return(pclist);
}
/* This method is to fill those 0 values from XCMS peaktable ( after "fill peaks" function from XCMS, some of the intensities are still 0 )
* This is done by generating 2 peaktables with file1 with "fillpeaks" function, file2 not, then sum all the "NA" areas in file1 and take average then assign to zero area.
* //*/
public static void Fill0Peaks(string fillPeaksDataPath, string noneFillPeaksDataPath, string outputFilePath)
{
List <string[]> fill0ValueList = new List <string[]>();
List <string> fillPeaksDataList = FileProcess.ReadFileToList(fillPeaksDataPath);
List <string> noneFillPeaksDataList = FileProcess.ReadFileToList(noneFillPeaksDataPath);
if (fillPeaksDataList.Count != noneFillPeaksDataList.Count)
{
Console.WriteLine("Fillpeaks and noneFillpeaks data are not match the same length! check the original data!");
return;
}
string[] titleLine_fill = null; string[] titleLine_nonefill = null;
if (fillPeaksDataList.Count != 0) // extract tittle row
{
titleLine_fill = fillPeaksDataList[0].Split(',');
titleLine_nonefill = noneFillPeaksDataList[0].Split(',');
}
double sumBackgroundNoise = 0, BackgroundNoise = 0;
int NAcount = 0;
for (int i = 1; i < noneFillPeaksDataList.Count; i++)
{
string[] line = noneFillPeaksDataList[i].Split(',');
for (int j = 0; j < line.GetLength(0); j++)
{
//if (titleLine[j].Replace("\"", "").StartsWith("X") &&!titleLine[j].Replace("\"", "").Contains("X.") && line[j] == "NA") {
if (titleLine_nonefill[j].Contains("_") && line[j] == "NA")
{
for (int k = 0; k < line.GetLength(0); k++)
{
if (titleLine_nonefill[j].Replace("\"", "") == titleLine_fill[k].Replace("\"", ""))
{
sumBackgroundNoise += double.Parse(fillPeaksDataList[i].Split(',')[k]);
NAcount++;
break;
}
}
}
}
}
if (NAcount != 0)
{
BackgroundNoise = sumBackgroundNoise / NAcount;
}
else
{
BackgroundNoise = 0;
}
for (int i = 0; i < fillPeaksDataList.Count; i++)
{
string[] line = fillPeaksDataList[i].Split(',');
for (int j = 0; j < line.GetLength(0); j++)
{
//if (titleLine[j].Replace("\"", "").StartsWith("X") && !titleLine[j].Replace("\"", "").Contains("X.") && line[j] == "0") {
if (titleLine_fill[j].Contains("_") && line[j] == "0")
{
line[j] = BackgroundNoise.ToString();
}
}
fill0ValueList.Add(line);
}
Console.WriteLine("BackgroundNoise: " + BackgroundNoise);
FileProcess.WritePeakList(fill0ValueList, outputFilePath);
}
static void Main(string[] args)
{
Stopwatch watch = new Stopwatch();
watch.Start();
string rootPath = @"E:\Different tissues\idMSMS different tissue\MIX Results\";
/**
* Fill 0 values in XCMS peaktable
*/
//string fillpeaksPath = rootPath + "NormPeakArea.csv";
//string nonefillpeaksPath = rootPath + "data_minfrac0.1_sn3_nonfill.csv";
//string outPath = rootPath + "NormPeakArea_fill0.csv";
//Data_Operation.Fill0Peaks(fillpeaksPath, nonefillpeaksPath, outPath);
/**
* export precursors (change parapeters to export all/without isotope/only adducts)
*/
//string CSVFilePath = rootPath + "NormPeakArea.csv";
//List<SpectrumData> allSpectraList = IdMSMS_Deconvolution.ProcessPeaktableToSpectralist(CSVFilePath);
//Data_Operation.FindPrecursorMZ(allSpectraList, outPrecursorPath);
/**
* Assemble idMS/MS with multi-eV
*/
string CSVFilePath = rootPath + "AllDeconvolutedSpectra_MS_pvals_9s.csv";
string outputFilePath = rootPath + "idmsms_3_6_0.9_minCor_rmv50_50ev.csv";
string NLOutputFilePath = rootPath + "NL_motif.csv";
string NDPmatrixPath = rootPath + "NDPmatrix.csv";
List <SpectrumData> allMergedList = IdMSMS_Deconvolution.MergeIdMSMSOfDiffCIDs(CSVFilePath);
FileProcess.WriteSpectrumList(allMergedList, outputFilePath); //output idMSMS spectra
FileProcess.OutputNeutralLoss(allMergedList, NLOutputFilePath); //output NL_motif
string commonNLPath = rootPath + "Common NL table.csv";
string outNLHitPath = rootPath + "NLmatrix.csv";
FileProcess.OutputSelectedNLto360Spec(allMergedList, commonNLPath, outNLHitPath); //output binary NL hit table
//List<SpectrumData> allSpectraList = FileProcess.ReadSpectralCSVFile(CSVFilePath); // use if there is spectra list to read
//List<SpectrumData> allKnownSpectraList = FileProcess.ReadKnownMSMSSpectraList(knownMSMSPath); // use if there is known spectra list
List <SpectrumData> allKnownSpectraList = allMergedList;
List <double[]> NDPResultList = new List <double[]>();
string[] queryTitleLine = new string[allMergedList.Count]; // add QueryTitle line to result
for (int i = 0; i < allMergedList.Count; i++)
{
queryTitleLine[i] = allMergedList[i].group.ToString();
}
string[] libraryTitleLine = new string[allKnownSpectraList.Count]; // add Library title to result
for (int i = 0; i < allKnownSpectraList.Count; i++)
{
libraryTitleLine[i] = allKnownSpectraList[i].peakList[0].sampleName;
}
for (int i = 0; i < allMergedList.Count; i++)
{
double[] line = new double[allMergedList.Count]; //
for (int j = 0; j < allMergedList.Count; j++) //
{
line[j] = Cal_Cosine_Product.CalCosineProduct(allMergedList[i], allMergedList[j]); //
}
NDPResultList.Add(line);
}
FileProcess.WriteNDPListToFile(NDPResultList, queryTitleLine, queryTitleLine, NDPmatrixPath); // output NDPmatrix
watch.Stop();
Console.WriteLine("weight: " + IdMSMS_Deconvolution.weight + " minSamp:" + IdMSMS_Deconvolution.minSampleNrForCor + " ndpThreh:" + IdMSMS_Deconvolution.ndpThreshold);
Console.WriteLine("Deconvoluted spectra number: " + allMergedList.Count);
Console.WriteLine("calculation time: " + watch.Elapsed);
Console.WriteLine("OK");
Console.Read();
}
/**
* After using this function，list contains a <PeakData> List， each index in the list coresponds to one mz （value averaged）
*/
public static List <SpectrumData> ProcessPeaktableToSpectralist(string csvFilePath) //read result CSV file generated by XCMS and CAMERA to a list of spectra
{
List <SpectrumData> allSpectrumList = new List <SpectrumData>();
List <PeakData> allPeakList = new List <PeakData>();
List <string> CSVList = new List <string>();
CSVList = FileProcess.ReadFileToList(csvFilePath);
//CSVList = BigFileReader.GetFileContent(csvFilePath);
string[] titleLine = null;
if (CSVList.Count != 0) // extract tittle row
{
titleLine = CSVList[0].Split(',');
}
else
{
Console.WriteLine("CSVFile is empty!");
}
int diffrtIndex = 0, pvalIndex = 0, featureIndex = 0, mzIndex = 4, rtIndex = 5,
isotopesIndex = titleLine.GetLength(0) - 4, adductIndex = titleLine.GetLength(0) - 3, pcgroupIndex = titleLine.GetLength(0) - 2, precursorIndex = titleLine.GetLength(0) - 1;
//store the index of interested tittles, allocate initial default values.
for (int j = 0; j < titleLine.GetLength(0); j++)
{
titleLine[j] = titleLine[j].Replace("\"", "");
switch (titleLine[j])
{
case "mz":
mzIndex = j;
break;
case "rt":
rtIndex = j;
break;
case "isotopes":
isotopesIndex = j;
break;
case "adduct":
adductIndex = j;
break;
case "pcgroup":
pcgroupIndex = j;
break;
case "pval":
pvalIndex = j;
break;
case "feature":
featureIndex = j;
break;
case "diffrt":
diffrtIndex = j;
break;
case "precursor":
precursorIndex = j;
break;
default:
break;
}
}
for (int i = 1; i < CSVList.Count; i++) // extract data rows, each time extract 1 row and take title row as a pair.
{
string[] line = CSVList[i].Replace("\"", "").Split(',');
List <PeakData> mzPeaks = new List <PeakData>();
for (int j = 0; j < line.GetLength(0); j++)
{
if (titleLine[j].Contains("_") && titleLine[j].Length > 1) // all the sample names contain "_" char
{
PeakData peak = new PeakData();
peak.mz = double.Parse(line[mzIndex]);
peak.rt = double.Parse(line[rtIndex]);
peak.isotopes = line[isotopesIndex];
peak.adduct = line[adductIndex];
peak.pcgroup = int.Parse(line[pcgroupIndex]);
peak.intensity = double.Parse(line[j]);
peak.sampleName = titleLine[j];
if (diffrtIndex != 0) //MS doesnt contain pval, feature, diffrt
{
peak.pval = double.Parse(line[pvalIndex]);
peak.feature = double.Parse(line[featureIndex]);
peak.diffrt = double.Parse(line[diffrtIndex]);
}
peak.precursor = line[precursorIndex].Replace(" ", "");
if (double.Parse(line[j]) >= weight * bg) //intensity which is 3 times higher than the average noise will be deemed as real peak!
{
peak.isNoise = false;
}
else
{
peak.isNoise = true;
}
mzPeaks.Add(peak);
}
}
PeakData averPeak = AverageIntensityofMZline(mzPeaks);
allPeakList.Add(averPeak);// after AverageIntensityofMZline, only one peakdata left
}
var groupResult = from p in allPeakList
group p by p.precursor;
foreach (var groups in groupResult)
{
List <PeakData> peakList = new List <PeakData>();
peakList = groups.ToList <PeakData>();
SpectrumData spec = new SpectrumData();
spec.peakList = peakList; //spec contains all original peak data
spec.group = groups.Key.Replace(" ", "");
if (peakList.Count != 0) //0 means no peaks, 1means single peak group, should be removed
{
allSpectrumList.Add(spec); //here merged all samples and CIDs into allSpectrumList
}
else
{
continue;
}
}
allSpectrumList = allSpectrumList.OrderBy(x => x.group).ToList();
return(allSpectrumList);
}
