private IEnumerable <RowKeyColKeyValue <string, string, UOPair <char> > > TripleEnumerable(HashSet <string> cidExcludeList, HashSet <string> snpExcludeSet)
{
CounterWithMessages counterWithMessages = new CounterWithMessages(SnpFile, compressionRatio: GZ ? .9 : 0);
using (TextReader textReader =
#if !SILVERLIGHT
GZ ? SnpFile.UnGZip() :
#endif
SnpFile.OpenText())
{
string line;
while (null != (line = textReader.ReadLine()))
{
counterWithMessages.Increment();
string[] field = line.Split('\t');
Helper.CheckCondition(field.Length == 4, "Expect lines of snp file to have four fields. " + line);
string snp = field[0];
string cid = field[1];
string value = field[2];
double confidence = double.Parse(field[3]);
Helper.CheckCondition(value.Length == 2 && value.All(c => "ACTG".Contains(c)), () => "Expect values in snp file to be a pair of ACT or G. " + value);
if (cidExcludeList.Contains(cid) || confidence < MissingThreshold || snpExcludeSet.Contains(snp))
{
continue; //not break;
}
yield return(RowKeyColKeyValue.Create(snp, cid, UOPair.Create(value[0], value[1])));
}
}
counterWithMessages.Finished();
}
private UOPair <char> StringToUOPairConverter(string pairString)
{
Helper.CheckCondition(pairString.Length == 3 && pairString[1] == ' ' && pairString[0] != '?' && pairString[2] != '?', "expect pair string in tped file to be three characters long with the middle being space and neither of the others being '?'. " + pairString);
if (pairString[0] == '0' || pairString[2] == '0')
{
Helper.CheckCondition(pairString[0] == '0' && pairString[2] == '0', "if either character in a pair string is '0' they should both be. " + pairString);
return(UOPair.Create('?', '?'));
}
return(UOPair.Create(pairString[0], pairString[2]));
}
private static IEnumerable <RowKeyColKeyValue <string, string, UOPair <char> > > TripleSequence(string genomePopFileName)
{
using (TextReader textReader = File.OpenText(genomePopFileName))
{
string firstLine = textReader.ReadLine(); //Ignore first line
Helper.CheckCondition(firstLine != null, "Expect genome pop file to contain at least one line");
string snpLine = textReader.ReadLine();
Helper.CheckCondition(snpLine != null, "Expect genome pop file to contain at least two lines");
string[] snpArray = snpLine.Split(',');
string line;
int cidIndex = -1;
while (null != (line = textReader.ReadLine()))
{
if (line == "pop")
{
continue; //not break
}
++cidIndex;
string cid = string.Format("cid{0}", cidIndex);
throw new Exception("Why did the next line have a ', StringSplitOptions.RemoveEmptyEntries'???");
string[] twoParts = line.Split(new char[] { ',' }, 2);
string[] valueArray = twoParts[1].TrimStart().Split(' ');
Helper.CheckCondition(valueArray.Length == snpArray.Length, "Expect each line to contain one entry per snp. " + cid);
for (int snpIndex = 0; snpIndex < snpArray.Length; ++snpIndex)
{
string value = valueArray[snpIndex];
string snp = snpArray[snpIndex];
switch (value)
{
case "0101":
yield return(RowKeyColKeyValue.Create(snp, cid, UOPair.Create('1', '1')));
break;
case "0102":
case "0201":
yield return(RowKeyColKeyValue.Create(snp, cid, UOPair.Create('1', '2')));
break;
case "0202":
yield return(RowKeyColKeyValue.Create(snp, cid, UOPair.Create('2', '2')));
break;
default:
throw new MatrixFormatException("Illegal value " + value);
}
}
}
}
}
/// <summary>
/// Returns true if and only if we know for sure that this genotype matches the given Hla. Returns false if an only if we
/// know for sure that this genotype does not describe the given hla. Otherwise, returns null.
/// </summary>
/// <param name="unambiguousHla"></param>
/// <returns></returns>
public bool?Matches(HlaI unambiguousHla, MixtureSemantics mixtureSemantics = MixtureSemantics.Uncertainty)
{
Helper.CheckCondition <ArgumentException>(!unambiguousHla.IsAmbiguous, "Can only check if you have an uynambiguous Hla");
Tuple <HlaI, HlaI> locus;
switch (unambiguousHla.Locus)
{
case HlaILocus.A:
locus = AAlleles; break;
case HlaILocus.B:
locus = BAlleles; break;
case HlaILocus.C:
locus = CAlleles; break;
default:
throw new Exception("Can't get here.");
}
UOPair <HlaI> locusToCompare = UOPair.Create(locus.Item1, locus.Item2); // easier to treat as unordered pair here, since it will order the nulls.
if (locusToCompare.First == null && locusToCompare.Second == null)
{
return(null); // we know nothing about this locus
}
// note: if one of them is null, it's first. so start with second.
bool?secondIsOther = locusToCompare.Second.Matches(unambiguousHla, mixtureSemantics);
if (secondIsOther.HasValue && secondIsOther.Value)
{
return(true); // if second is a match, then we know it has it.
}
if (locusToCompare.First == null)
{
return(null); //don't know anything about first.
}
bool?firstIsOther = locusToCompare.First.Matches(unambiguousHla, mixtureSemantics);
if (firstIsOther.HasValue && firstIsOther.Value)
{
return(true); // first is a match, so we know we have it
}
if (!firstIsOther.HasValue || !secondIsOther.HasValue)
{
return(null); // neither has it, so if either is missing, we don't know
}
return(false); // if we get here, then both alleles reported false.
}
private IEnumerable <RowKeyColKeyValue <string, string, UOPair <char> > > TripleEnumerable()
{
CounterWithMessages counterWithMessages = new CounterWithMessages("Reading " + SnpFile.Name, messageIntervalOrNull: 1000);
using (TextReader textReader = SnpFile.OpenText())
{
string headerLine = textReader.ReadLine();
Helper.CheckCondition(headerLine != null, "Expect file to contain a first line");
string[] headerFields = headerLine.Split('\t');
Helper.CheckCondition(headerFields.Length > 0 && headerFields[0] == "", "Expect first column of first line to be blank");
string line;
while (null != (line = textReader.ReadLine()))
{
counterWithMessages.Increment();
string[] fields = line.Split('\t');
Helper.CheckCondition(fields.Length == headerFields.Length, "Expect all lines to have the same # of columns");
string cid = fields[0];
for (int snpIndex = 1; snpIndex < headerFields.Length; ++snpIndex) // start at one to skip over 1st column
{
string snp = headerFields[snpIndex];
string valueInFile = fields[snpIndex];
UOPair <char> uoPair;
if (valueInFile == "-")
{
continue; // not break;
}
else if (valueInFile.Length == 1)
{
char c = valueInFile[0];
Helper.CheckCondition("ACTG".Contains(c), () => "Expect values in snp file to be ACT or G. " + valueInFile);
uoPair = UOPair.Create(c, c);
}
else
{
Helper.CheckCondition(valueInFile.Length == 3 && valueInFile[1] == '/' && "ACTG".Contains(valueInFile[0]) && "ACTG".Contains(valueInFile[2]), () => "Expect longer values in snp file be of the form 'a/b' where a & b are ACT or G");
uoPair = UOPair.Create(valueInFile[0], valueInFile[2]);
}
yield return(RowKeyColKeyValue.Create(snp, cid, uoPair));
}
}
}
}
private IEnumerable <RowKeyColKeyValue <string, string, UOPair <char> > > TripleEnumerable()
{
//int? totalLineCountOrNull = null;
//int? messageIntervalOrNull = 10000;
//using (TextReader textReader = SnpFile.OpenText())
//{
// string line = textReader.ReadLine();
// if (null != line || line.Length == 0)
// {
// totalLineCountOrNull = (int?)(SnpFile.Length / (long)(line.Length + 2 /*line end*/));
// messageIntervalOrNull = null;
// }
//}
CounterWithMessages counterWithMessages = new CounterWithMessages(SnpFile); //"Reading " + SnpFile.Name, messageIntervalOrNull, totalLineCountOrNull);
using (TextReader textReader = SnpFile.OpenText())
{
string line;
while (null != (line = textReader.ReadLine()))
{
counterWithMessages.Increment();
string[] field = line.Split('\t');
Helper.CheckCondition(field.Length == 3, "Expect lines of snp file to have three fields. " + line);
string cid = field[0];
string snp = field[1];
string value = field[2];
if (value == "00")
{
continue; //not break;
}
Helper.CheckCondition(value.Length == 2 && value.All(c => "ACTG".Contains(c)), () => "Expect values in snp file to be a pair of ACTG or 00. " + value);
yield return(RowKeyColKeyValue.Create(snp, cid, UOPair.Create(value[0], value[1])));
}
counterWithMessages.Finished();
}
}
public StringToUOPairConverter() : base(s => UOPair.Create(s[0], s[1]), pair => string.Format("{0}{1}", pair.First, pair.Second))
{
}