/// <summary>
/// Load known CN data from a .bed file. File lines have fields:
/// chromosome, start, end, chromcountA, chromcountB
/// So, copy number is the sum of the last 2 fields, major chromosome count is the max of the last 2 fields.
/// </summary>
/// <param name="oracleBedPath"></param>
protected Dictionary<string, List<CNInterval>> LoadIntervalsFromBed(string oracleBedPath, bool getCN)
{
bool stripChr = false;
int count = 0;
long totalBases = 0;
Dictionary<string, List<CNInterval>> bedIntervals = new Dictionary<string, List<CNInterval>>();
using (StreamReader reader = new StreamReader(oracleBedPath))
{
while (true)
{
string fileLine = reader.ReadLine();
if (fileLine == null) break;
if (fileLine.Length == 0 || fileLine[0] == '#') continue;
string[] bits = fileLine.TrimEnd('\t').Split('\t');
if (bits.Length < 3) continue;
string chromosome = bits[0];
if (stripChr) chromosome = chromosome.Replace("chr", "");
if (!bedIntervals.ContainsKey(chromosome)) bedIntervals[chromosome] = new List<CNInterval>();
CNInterval interval = new CNInterval();
interval.Start = int.Parse(bits[1]);
interval.End = int.Parse(bits[2]);
if (getCN) // bits.Length >= 5)
{
interval.CN = int.Parse(bits[3]) + int.Parse(bits[4]);
}
totalBases += interval.Length;
bedIntervals[chromosome].Add(interval);
count++;
}
}
Console.WriteLine(">>>Loaded {0} CN intervals ({1} bases)", count, totalBases);
return bedIntervals;
}
/// <summary>
/// Load known CN data from a .bed file. File lines have fields:
/// chromosome, start, end, chromcountA, chromcountB
/// So, copy number is the sum of the last 2 fields, major chromosome count is the max of the last 2 fields.
/// </summary>
/// <param name="oracleBedPath"></param>
/// <param name="getCn"></param>
/// <param name="heterogeneityFraction"></param>
protected static Dictionary <string, List <CNInterval> > LoadIntervalsFromBed(string oracleBedPath, bool getCn, double heterogeneityFraction)
{
bool stripChr = false;
int count = 0;
long totalBases = 0;
Dictionary <string, List <CNInterval> > bedIntervals = new Dictionary <string, List <CNInterval> >();
using (FileStream stream = new FileStream(oracleBedPath, FileMode.Open, FileAccess.Read))
using (StreamReader reader = new StreamReader(stream))
{
while (true)
{
string fileLine = reader.ReadLine();
if (fileLine == null)
{
break;
}
if (fileLine.Length == 0 || fileLine[0] == '#')
{
continue;
}
string[] bits = fileLine.TrimEnd('\t').Split('\t');
if (bits.Length < 3)
{
continue;
}
string chromosome = bits[0];
if (stripChr)
{
chromosome = chromosome.Replace("chr", "");
}
if (!bedIntervals.ContainsKey(chromosome))
{
bedIntervals[chromosome] = new List <CNInterval>();
}
CNInterval interval = new CNInterval(chromosome);
interval.Start = int.Parse(bits[1]);
interval.End = int.Parse(bits[2]);
if (getCn) // bits.Length >= 5)
{
if (heterogeneityFraction < 1 && bits.Length > 5 && int.Parse(bits[3]) == 1 && int.Parse(bits[4]) == 1)
{
if (heterogeneityFraction > double.Parse(bits[5]))
{
continue;
}
}
interval.Cn = int.Parse(bits[3]) + int.Parse(bits[4]);
}
totalBases += interval.Length;
bedIntervals[chromosome].Add(interval);
count++;
}
}
Console.WriteLine(">>>Loaded {0} CN intervals ({1} bases)", count, totalBases);
return(bedIntervals);
}
private static CNInterval ParseCnInterval(string fileLine)
{
string[] bits = fileLine.Split('\t');
string chromosome = bits[0];
CNInterval interval = new CNInterval(chromosome)
{
Start = int.Parse(bits[1]),
Cn = -1
};
string[] infoBits = bits[7].Split(';');
foreach (string subBit in infoBits)
{
if (subBit.StartsWith("CN="))
{
float tempCn = float.Parse(subBit.Substring(3));
if (subBit.EndsWith(".5"))
{
interval.Cn = (int)Math.Round(tempCn + 0.1); // round X.5 up to X+1
}
else
{
interval.Cn = (int)Math.Round(tempCn); // Round off
}
}
if (subBit.StartsWith("END="))
{
interval.End = int.Parse(subBit.Substring(4));
}
}
// Parse CN from Canvas output:
if (bits.Length > 8)
{
string[] subBits = bits[8].Split(':');
string[] subBits2 = bits[9].Split(':');
for (int subBitIndex = 0; subBitIndex < subBits.Length; subBitIndex++)
{
if (subBits[subBitIndex] == "CN")
{
interval.Cn = int.Parse(subBits2[subBitIndex]);
}
}
}
if (interval.End == 0 || interval.Cn < 0)
{
throw new ArgumentException("Invalid record. End cannot be 0 and CN must be >= 0");
}
return(interval);
}
protected void LoadKnownCNVCF(string oracleVCFPath)
{
bool stripChr = false;
// Load our "oracle" of known copy numbers:
this.KnownCN = new Dictionary<string, List<CNInterval>>();
int count = 0;
using (GzipReader reader = new GzipReader(oracleVCFPath))
{
while (true)
{
string fileLine = reader.ReadLine();
if (fileLine == null) break;
if (fileLine.Length == 0 || fileLine[0] == '#') continue;
string[] bits = fileLine.Split('\t');
string chromosome = bits[0];
if (stripChr) chromosome = chromosome.Replace("chr", "");
if (!KnownCN.ContainsKey(chromosome)) KnownCN[chromosome] = new List<CNInterval>();
CNInterval interval = new CNInterval();
interval.Start = int.Parse(bits[1]);
interval.CN = -1;
string[] infoBits = bits[7].Split(';');
foreach (string subBit in infoBits)
{
if (subBit.StartsWith("CN="))
{
float tempCN = float.Parse(subBit.Substring(3));
if (subBit.EndsWith(".5"))
{
interval.CN = (int)Math.Round(tempCN + 0.1); // round X.5 up to X+1
}
else
{
interval.CN = (int)Math.Round(tempCN); // Round off
}
}
if (subBit.StartsWith("END="))
{
interval.End = int.Parse(subBit.Substring(4));
}
}
// Parse CN from Canvas output:
if (bits.Length > 8)
{
string[] subBits = bits[8].Split(':');
string[] subBits2 = bits[9].Split(':');
for (int subBitIndex = 0; subBitIndex < subBits.Length; subBitIndex++)
{
if (subBits[subBitIndex] == "CN")
{
interval.CN = int.Parse(subBits2[subBitIndex]);
}
}
}
if (interval.End == 0 || interval.CN < 0)
{
Console.WriteLine("Error - bogus record!");
Console.WriteLine(fileLine);
}
else
{
KnownCN[chromosome].Add(interval);
count++;
}
}
}
Console.WriteLine(">>>Loaded {0} known-CN intervals", count);
}
protected void LoadKnownCNVCF(string oracleVCFPath)
{
bool stripChr = false;
// Load our "oracle" of known copy numbers:
this.KnownCN = new Dictionary <string, List <CNInterval> >();
int count = 0;
using (GzipReader reader = new GzipReader(oracleVCFPath))
{
while (true)
{
string fileLine = reader.ReadLine();
if (fileLine == null)
{
break;
}
if (fileLine.Length == 0 || fileLine[0] == '#')
{
continue;
}
string[] bits = fileLine.Split('\t');
string chromosome = bits[0];
if (stripChr)
{
chromosome = chromosome.Replace("chr", "");
}
if (!KnownCN.ContainsKey(chromosome))
{
KnownCN[chromosome] = new List <CNInterval>();
}
CNInterval interval = new CNInterval();
interval.Start = int.Parse(bits[1]);
interval.CN = -1;
string[] infoBits = bits[7].Split(';');
foreach (string subBit in infoBits)
{
if (subBit.StartsWith("CN="))
{
float tempCN = float.Parse(subBit.Substring(3));
if (subBit.EndsWith(".5"))
{
interval.CN = (int)Math.Round(tempCN + 0.1); // round X.5 up to X+1
}
else
{
interval.CN = (int)Math.Round(tempCN); // Round off
}
}
if (subBit.StartsWith("END="))
{
interval.End = int.Parse(subBit.Substring(4));
}
}
// Parse CN from Canvas output:
if (bits.Length > 8)
{
string[] subBits = bits[8].Split(':');
string[] subBits2 = bits[9].Split(':');
for (int subBitIndex = 0; subBitIndex < subBits.Length; subBitIndex++)
{
if (subBits[subBitIndex] == "CN")
{
interval.CN = int.Parse(subBits2[subBitIndex]);
}
}
}
if (interval.End == 0 || interval.CN < 0)
{
Console.WriteLine("Error - bogus record!");
Console.WriteLine(fileLine);
}
else
{
KnownCN[chromosome].Add(interval);
count++;
}
}
}
Console.WriteLine(">>>Loaded {0} known-CN intervals", count);
}
