public static List <GenomicBin> ReadFromTextFile(string infile)
{
List <GenomicBin> bins = new List <GenomicBin>();
using (GzipReader reader = new GzipReader(infile))
{
string row;
while ((row = reader.ReadLine()) != null)
{
string[] fields = row.Split('\t');
string chr = fields[0];
int start = Convert.ToInt32(fields[1]);
int stop = Convert.ToInt32(fields[2]);
//int count = Convert.ToInt32(fields[3]);
float count = float.Parse(fields[3]);
int gc = Convert.ToInt32(fields[4]);
GenomicBin bin = new GenomicBin(chr, start, stop, gc, count);
bins.Add(bin);
}
}
return(bins);
}
public SampleGenomicBin(string chr, int start, int stop, float count)
{
_genomicBin = new GenomicBin();
GenomicBin.Chromosome = chr;
GenomicBin.Interval = new BedInterval(start, stop);
this.Count = count;
}
public static List<GenomicBin> ReadFromTextFile(string infile)
{
List<GenomicBin> bins = new List<GenomicBin>();
using (GzipReader reader = new GzipReader(infile))
{
string row;
while ((row = reader.ReadLine()) != null)
{
string[] fields = row.Split('\t');
string chr = fields[0];
int start = Convert.ToInt32(fields[1]);
int stop = Convert.ToInt32(fields[2]);
//int count = Convert.ToInt32(fields[3]);
float count = float.Parse(fields[3]);
int gc = Convert.ToInt32(fields[4]);
GenomicBin bin = new GenomicBin(chr, start, stop, gc, count);
bins.Add(bin);
}
}
return bins;
}
public SampleGenomicBin(string chr, int start, int stop, int gc)
{
_genomicBin = new GenomicBin();
GenomicBin.Chromosome = chr;
GenomicBin.Interval = new BedInterval(start, stop);
GenomicBin.GC = gc;
this.CountDeviation = -1;
}
private static Tuple <int, int> GetEndConfidenceInterval(GenomicBin thisBin, GenomicBin nextBin)
{
if (nextBin == null || thisBin.Interval.End != nextBin.Interval.Start)
{
return(GetConfidenceInterval(thisBin));
}
return(Tuple.Create(-GetHalfLength(thisBin), GetHalfLength(nextBin)));
}
private static Tuple <int, int> GetStartConfidenceInterval(GenomicBin thisBin, GenomicBin previousBin)
{
if (previousBin == null || previousBin.Interval.End != thisBin.Interval.Start)
{
return(GetConfidenceInterval(thisBin));
}
return(Tuple.Create(-GetHalfLength(previousBin), GetHalfLength(thisBin)));
}
public SampleGenomicBin()
{
_genomicBin = new GenomicBin();
GenomicBin.Chromosome = null;
GenomicBin.Interval = new GenomicInterval();
GenomicBin.GC = -1;
this.CountDeviation = -1;
}
public SampleGenomicBin(string chr, int start, int stop, int gc, float count, double MadOfDIffs)
{
_genomicBin = new GenomicBin();
GenomicBin.Chromosome = chr;
GenomicBin.Interval = new BedInterval(start, stop);
GenomicBin.GC = gc;
this.Count = count;
this.CountDeviation = MadOfDIffs;
}
public SampleGenomicBin(string chr, int start, int stop, int gc, float count)
{
_genomicBin = new GenomicBin(chr, new GenomicInterval()
{
Start = start, End = stop
}, gc);
this.Count = count;
this.CountDeviation = -1;
}
public static Dictionary <string, List <GenomicBin> > LoadBedFile(string bedPath, int?gcIndex = null)
{
Dictionary <string, List <GenomicBin> > excludedIntervals = new Dictionary <string, List <GenomicBin> >();
int count = 0;
using (StreamReader reader = new StreamReader(bedPath))
{
while (true)
{
string fileLine = reader.ReadLine();
if (fileLine == null)
{
break;
}
string[] bits = fileLine.Split('\t');
string chr = bits[0];
if (!excludedIntervals.ContainsKey(chr))
{
excludedIntervals[chr] = new List <GenomicBin>();
}
GenomicBin interval = new GenomicBin();
interval.Chromosome = chr;
interval.Start = int.Parse(bits[1]);
interval.Stop = int.Parse(bits[2]);
if (interval.Start < 0)
{
throw new ApplicationException(String.Format("Start must be non-negative in a BED file: {0}", fileLine));
}
if (interval.Start >= interval.Stop) // Do not allow empty intervals
{
throw new ApplicationException(String.Format("Start must be less than Stop in a BED file: {0}", fileLine));
}
if (gcIndex.HasValue && gcIndex.Value < bits.Length)
{
interval.GC = int.Parse(bits[gcIndex.Value]);
}
excludedIntervals[chr].Add(interval);
count++;
}
}
Console.WriteLine(">>> Loaded {0} intervals for {1} sequences", count, excludedIntervals.Keys.Count);
return(excludedIntervals);
}
private static Tuple <int, int> GetConfidenceInterval(GenomicBin bin)
{
var halfLength = GetHalfLength(bin);
return(Tuple.Create(-halfLength, halfLength));
}
public void TestBins()
{
GenomicBin bin = new GenomicBin("chr1", 12345, 678910, 20, 100);
Assert.AreEqual(bin.Size, 666565);
}
private static int GetHalfLength(GenomicBin bin)
{
return((int)Math.Round(bin.Interval.Length / 2.0, MidpointRounding.AwayFromZero));
}
public SampleGenomicBin(string chr, int start, int stop, int gc, float count)
{
_genomicBin = new GenomicBin(chr, new BedInterval(start, stop), gc);
this.Count = count;
this.CountDeviation = -1;
}
/// <summary>
/// Populate the list of GenomicBin objects for this chromosome.
/// </summary>
static void BinCountsForChromosome(BinTaskArguments arguments)
{
List<GenomicBin> bins = arguments.Bins;
bool usePredefinedBins = bins.Any();
int predefinedBinIndex = 0;
GenericRead fastaEntry = arguments.FastaEntry; //fastaEntryKVP.Value;
BinState currentBin = new BinState();
string chr = arguments.Chromosome;
BitArray possibleAlignments = arguments.PossibleAlignments;
HitArray observedAlignments = arguments.ObservedAlignments;
CanvasCoverageMode coverageMode = arguments.CoverageMode;
int pos = usePredefinedBins ? bins[predefinedBinIndex].Start : 0;
// Skip past leading Ns
while (fastaEntry.Bases[pos].Equals('n'))
pos++;
List<float> binPositions = new List<float>();
List<int> binObservations = new List<int>();
for (; pos < fastaEntry.Bases.Length; pos++)
{
// Sets the start of the bin
if (currentBin.StartPosition == -1)
currentBin.StartPosition = pos;
if (!fastaEntry.Bases[pos].Equals("n"))
currentBin.NucleotideCount++;
//if (IsGC(fastaEntry.Bases[pos]))
// currentBin.GCCount++;
switch (fastaEntry.Bases[pos])
{
case 'C':
case 'c':
case 'G':
case 'g':
currentBin.GCCount++;
break;
}
if (possibleAlignments[pos])
{
currentBin.PossibleCount++;
currentBin.ObservedCount += observedAlignments.Data[pos];
binObservations.Add(observedAlignments.Data[pos]);
if (coverageMode == CanvasCoverageMode.GCContentWeighted)
binPositions.Add(arguments.ObservedVsExpectedGC[arguments.ReadGCContent[pos]]);
}
// We've seen the desired number of possible alignment positions.
if ((!usePredefinedBins && currentBin.PossibleCount == arguments.BinSize)
|| (usePredefinedBins && pos == bins[predefinedBinIndex].Stop - 1))
{
if (coverageMode == CanvasCoverageMode.TruncatedDynamicRange) // Truncated dynamic range
{
currentBin.ObservedCount = 0;
foreach (int Value in binObservations)
{
currentBin.ObservedCount += Math.Min(10, Value);
}
}
if (coverageMode == CanvasCoverageMode.GCContentWeighted) // read GC content weighted
{
currentBin.ObservedCount = 0;
float tmpObservedCount = 0;
for (int i = 0; i < binObservations.Count; i++)
{
tmpObservedCount += Math.Min(10, (float)binObservations[i] / binPositions[i]);
}
currentBin.ObservedCount = (int)Math.Round(tmpObservedCount);
}
int gc = (int)(100 * currentBin.GCCount / currentBin.NucleotideCount);
if (usePredefinedBins)
{
bins[predefinedBinIndex].GC = gc;
bins[predefinedBinIndex].Count = currentBin.ObservedCount;
predefinedBinIndex++;
if (predefinedBinIndex >= bins.Count) { break; } // we have processed all the bins
pos = bins[predefinedBinIndex].Start - 1; // jump to right before the next predefined bin
}
else
{
// Note the pos + 1 to make the first three conform to bed specification
GenomicBin bin = new GenomicBin(chr, currentBin.StartPosition, pos + 1, gc, currentBin.ObservedCount);
bins.Add(bin);
}
// Reset all relevant variables
currentBin.Reset();
binObservations.Clear();
binPositions.Clear();
}
}
}
/// <summary>
/// Read predefined bins from a BED file. Assume the bins are sorted by genomic coordinates.
/// </summary>
/// <param name="predefinedBinsFile">input BED file</param>
/// <returns>predefined bins by chromosome</returns>
static Dictionary<string, List<GenomicBin>> ReadPredefinedBins(string predefinedBinsFile)
{
Dictionary<string, List<GenomicBin>> predefinedBins = new Dictionary<string, List<GenomicBin>>();
if (!File.Exists(predefinedBinsFile)) { return predefinedBins; }
using (StreamReader reader = new StreamReader(predefinedBinsFile))
{
string row;
while ((row = reader.ReadLine()) != null)
{
try
{
if (row.StartsWith("#")) { continue; } // ignore comments
string[] fields = row.Split('\t');
if (fields.Length < 3) { continue; }
string chr = fields[0];
int start = Convert.ToInt32(fields[1]);
int stop = Convert.ToInt32(fields[2]);
GenomicBin bin = new GenomicBin(chr, start, stop, 0, 0);
if (!predefinedBins.ContainsKey(chr)) { predefinedBins[chr] = new List<GenomicBin>(); }
predefinedBins[chr].Add(bin);
}
catch (Exception e)
{
throw new Exception(String.Format("Failed to parse {0}; Line: {1}", predefinedBinsFile, row), e);
}
}
}
return predefinedBins;
}
public MultiSampleGenomicBin(GenomicBin genomicBin, List <float> counts)
{
Bin = genomicBin;
Counts = counts;
}
static public Dictionary<string, List<GenomicBin>> LoadBedFile(string bedPath)
{
Dictionary<string, List<GenomicBin>> excludedIntervals = new Dictionary<string, List<GenomicBin>>();
int count = 0;
using (StreamReader reader = new StreamReader(bedPath))
{
while (true)
{
string fileLine = reader.ReadLine();
if (fileLine == null) break;
string[] bits = fileLine.Split('\t');
string chr = bits[0];
if (!excludedIntervals.ContainsKey(chr)) excludedIntervals[chr] = new List<GenomicBin>();
GenomicBin interval = new GenomicBin();
interval.Start = int.Parse(bits[1]);
interval.Stop = int.Parse(bits[2]);
excludedIntervals[chr].Add(interval);
count++;
}
}
Console.WriteLine(">>> Loaded {0} excluded intervals for {1} sequences", count, excludedIntervals.Keys.Count);
return excludedIntervals;
}
/// <summary>
/// Loads .cleaned bed files, merges bins from multiple samples and returns GenomicBin objects with MultiSampleCount
/// </summary>
public static Dictionary <string, List <MultiSampleGenomicBin> > MergeMultiSampleCleanedBedFile(List <IFileLocation> canvasCleanBedPaths)
{
// initialize variables to hold multi-sample bed files
Dictionary <string, List <MultiSampleGenomicBin> > multiSampleGenomicBins = new Dictionary <string, List <MultiSampleGenomicBin> >();
Dictionary <string, Dictionary <int, int> > start = new Dictionary <string, Dictionary <int, int> >();
Dictionary <string, Dictionary <int, int> > stop = new Dictionary <string, Dictionary <int, int> >();
Dictionary <string, Dictionary <int, List <float> > > binCounts = new Dictionary <string, Dictionary <int, List <float> > >();
List <int> counts = new List <int>();
HashSet <string> chromosomes = new HashSet <string>();
Console.WriteLine("Merge and normalize CanvasClean bed files");
foreach (IFileLocation bedPath in canvasCleanBedPaths)
{
int count = 0;
using (GzipReader reader = new GzipReader(bedPath.FullName))
{
while (true)
{
string fileLine = reader.ReadLine();
if (fileLine == null)
{
break;
}
string[] lineBedFile = fileLine.Split('\t');
string chr = lineBedFile[0];
if (!chromosomes.Contains(chr))
{
chromosomes.Add(chr);
}
count++;
}
}
counts.Add(count);
Console.WriteLine($"count {count}");
}
foreach (string chr in chromosomes)
{
start[chr] = new Dictionary <int, int>();
stop[chr] = new Dictionary <int, int>();
binCounts[chr] = new Dictionary <int, List <float> >();
}
// read counts and segmentIDs
foreach (IFileLocation bedPath in canvasCleanBedPaths)
{
Console.WriteLine(bedPath);
using (GzipReader reader = new GzipReader(bedPath.FullName))
{
while (true)
{
string fileLine = reader.ReadLine();
if (fileLine == null)
{
break;
}
string[] lineBedFile = fileLine.Split('\t');
string chr = lineBedFile[0];
int pos = int.Parse(lineBedFile[1]);
start[chr][pos] = pos;
stop[chr][pos] = int.Parse(lineBedFile[2]);
if (binCounts[chr].ContainsKey(pos))
{
binCounts[chr][pos].Add(float.Parse(lineBedFile[3]));
}
else
{
binCounts[chr][pos] = new List <float> {
float.Parse(lineBedFile[3])
}
};
}
}
}
Console.WriteLine("create GenomeBin intervals");
// create GenomeBin intervals
foreach (string chr in chromosomes)
{
if (!multiSampleGenomicBins.ContainsKey(chr))
{
multiSampleGenomicBins[chr] = new List <MultiSampleGenomicBin>();
}
var binStartPositions = start[chr].Keys.ToList();
foreach (var binStartPosition in binStartPositions)
{
// if outlier is removed in one sample, remove it in all samples
if (binCounts[chr][binStartPosition].Count < canvasCleanBedPaths.Count)
{
continue;
}
if (binStartPosition < 0)
{
throw new Illumina.Common.IlluminaException($"Start must be non-negative");
}
if (binStartPosition >= stop[chr][binStartPosition]) // Do not allow empty intervals
{
throw new Illumina.Common.IlluminaException($"Start must be less than Stop");
}
GenomicBin interval = new GenomicBin(chr, new BedInterval(binStartPosition, stop[chr][binStartPosition]));
multiSampleGenomicBins[chr].Add(new MultiSampleGenomicBin(interval, binCounts[chr][binStartPosition]));
}
}
return(multiSampleGenomicBins);
}