public void SplitOnProtein(string niceName, string outputDirectory)
{
Debug.WriteLine(niceName);
Dictionary <string, Dictionary <string, string> > proteinToCaseIdToSequence = new Dictionary <string, Dictionary <string, string> >();
foreach (string caseId in _caseIdToAASeq.Keys)
{
AASeq aaSeq = _caseIdToAASeq[caseId];
string previousProtein = null;
Dictionary <string, StringBuilder> proteinToSequence = new Dictionary <string, StringBuilder>();
for (int aa0Pos = 0; aa0Pos < (int)SequenceLengthOrNull; ++aa0Pos)
{
string posName = aaSeq.OriginalAA1Position(aa0Pos);
string[] posParts = posName.Split('@');
string protein = posParts[0];
StringBuilder sequence = proteinToSequence.GetValueOrDefault(protein);
if (previousProtein != protein)
{
Helper.CheckCondition(sequence.Length == 0, "Expect proteins to be contintiguous");
previousProtein = protein;
}
Set <char> strainAASet = aaSeq[aa0Pos];
sequence.Append(AASeq.AaAsString(strainAASet));
}
foreach (string protein in proteinToSequence.Keys)
{
Dictionary <string, string> caseIdToSequence = proteinToCaseIdToSequence.GetValueOrDefault(protein);
caseIdToSequence.Add(caseId, proteinToSequence[protein].ToString());
}
}
foreach (string protein in proteinToCaseIdToSequence.Keys)
{
string outputFileName = string.Format(@"{0}\{1}.{2}.aaSeq.txt", outputDirectory, protein, niceName);
using (TextWriter textWriter = File.CreateText(outputFileName))
{
textWriter.WriteLine("cid\taaSeq"); //!!!const
Dictionary <string, string> caseIdToSequence = proteinToCaseIdToSequence[protein];
foreach (string caseId in caseIdToSequence.Keys)
{
textWriter.WriteLine(Helper.CreateTabString(caseId, caseIdToSequence[caseId]));
}
}
}
}
public void TriplesAppend(ref Set <string> seenTriple, ref Dictionary <string, List <string> > proteinToTripleList)
{
AASeq aaSeq = GetFirstAASeq();
for (int aa0Pos = 0; aa0Pos < (int)SequenceLengthOrNull; ++aa0Pos)
{
string posName = aaSeq.OriginalAA1Position(aa0Pos);
string[] posParts = posName.Split('@');
string protein = posParts[0];
string hxb2Pos = posParts[2];
string triple = Helper.CreateTabString(protein, hxb2Pos);
if (!seenTriple.Contains(triple))
{
seenTriple.AddNew(triple);
List <string> tripleList = proteinToTripleList.GetValueOrDefault(protein);
tripleList.Add(triple);
}
}
}
static public AASeq GetCompressedInstance(string caseId, AASeq aaSeqIn, bool stopOnStop, TextWriter errorStream)
{
AASeq aaSeqOut = new AASeq(aaSeqIn.MixtureSemantics);
aaSeqOut.Sequence = new List <Set <char> >();
aaSeqOut._originalAA1PositionTableOrNull = new List <string>();
Debug.Assert(aaSeqOut.Offset == 0); // real assert
for (int iChar = 0; iChar < aaSeqIn.Count; ++iChar)
{
Set <char> set = aaSeqIn[iChar];
string originalAA1Position = aaSeqIn.OriginalAA1Position(iChar);
if (set.Equals(Delete)) //!!!const
{
continue;
}
if (set.Equals(Stop)) //!!!const
{
if (iChar != aaSeqIn.Count - 1)
{
errorStream.WriteLine("Warning: The sequence for case id '{0}' contains a '*' before the last position", caseId);
if (stopOnStop)
{
break;
}
}
else
{
break;
}
}
aaSeqOut.Sequence.Add(set);
aaSeqOut._originalAA1PositionTableOrNull.Add(originalAA1Position);
}
return(aaSeqOut);
}
public IEnumerable <string> SparseLineEnumeration(bool keepOneValueVariables)
{
if (_caseIdToAASeq.Count == 0)
{
Debug.Assert(SequenceLengthOrNull == null); // real assert
yield break;
}
Helper.CheckCondition(SequenceLengthOrNull != null, "This converter to sparse assumes all sequences have the same length");
/*
* n1pos aa pid val
* 880 A 3 F
* 880 A 5 F
* 880 A 9 F
* 880 A 13 F
* 880 A 14 F
* 880 A 15 T
* ...
*/
for (int aa0Pos = 0; aa0Pos < (int)SequenceLengthOrNull; ++aa0Pos)
{
Set <char> everyAminoAcid = EveryAminoAcid(aa0Pos);
if (!keepOneValueVariables && everyAminoAcid.Count == 1)
{
continue;
}
string posName = null;
foreach (char aa in everyAminoAcid)
{
Set <bool> valueSet = Set <bool> .GetInstance();
Dictionary <string, bool> caseToVal = new Dictionary <string, bool>();
foreach (string caseId in _caseIdToAASeq.Keys)
{
AASeq aaSeq = _caseIdToAASeq[caseId];
if (aa0Pos >= aaSeq.Count)
{
continue;
}
//Helper.CheckCondition(aaSeq.IsUsingOriginalPositions(), "This converter to sparse assumes all sequences are using their original positions");
Set <char> strainAASet = aaSeq[aa0Pos];
if (posName == null)
{
posName = aaSeq.OriginalAA1Position(aa0Pos);
//if (posName.Contains("68.3B"))
//{
// Console.WriteLine("Found it first");
//}
}
else
{
Helper.CheckCondition(posName == aaSeq.OriginalAA1Position(aa0Pos));
}
// missing: e.g. A/Any or A/AB
// 1: e.g. A/A
// 0: e.g. A/B or A/BCD
if (strainAASet.Equals(AASeq.Any))
{
//Do nothing - missing
}
else if (strainAASet.Contains(aa))
{
if (strainAASet.Count > 1)
{
switch (aaSeq.MixtureSemantics)
{
case MixtureSemantics.Pure:
caseToVal.Add(caseId, false);
valueSet.AddNewOrOld(false);
break;
case MixtureSemantics.Uncertainty:
// Do nothing = missing
break;
case MixtureSemantics.Any:
caseToVal.Add(caseId, true);
valueSet.AddNewOrOld(true);
break;
default:
Helper.CheckCondition(false, "Unknown mixturesemantics " + aaSeq.MixtureSemantics.ToString());
break;
}
}
else
{
caseToVal.Add(caseId, true);
valueSet.AddNewOrOld(true);
}
}
else
{
caseToVal.Add(caseId, false);
valueSet.AddNewOrOld(false);
}
}
Helper.CheckCondition(posName != null);
if (keepOneValueVariables || valueSet.Count == 2)
{
foreach (KeyValuePair <string, bool> caseIdAndVal in caseToVal)
{
string variableName = string.Format("{0}@{1}", posName, aa);
//string variableName = string.Format("{1}@{0}", posName, aa);
//if (variableName.Contains("68.3B"))
//{
// Console.WriteLine("Found it first");
//}
yield return(Helper.CreateTabString(
variableName, caseIdAndVal.Key, caseIdAndVal.Value ? 1 : 0));
}
}
}
}
}