public static NamedSequence GetConsensus(List <NamedSequence> seqs)
{
int len = -1;
//StringBuilder consensusSeq = new StringBuilder();
List <AASeq> aaSeqs = new List <AASeq>();
bool isDna = seqs[0].IsDna();
foreach (NamedSequence seq in seqs)
{
AASeq aaSeq = isDna ? DnaSeq.GetInstance(seq.Sequence, MixtureSemantics.Uncertainty) : AASeq.GetInstance(seq.Sequence, MixtureSemantics.Uncertainty);
len = Math.Max(len, aaSeq.Count);
//if (len < 0)
//{
// len = aaSeq.Count;
//}
//else
//{
// Helper.CheckCondition(len == aaSeq.Count, String.Format("Sequence {0} is a different length from previous sequences", seq.Name));
//}
aaSeqs.Add(aaSeq);
}
//for (int i = 0; i < len; i++)
//{
// Dictionary<string, int> charToCount = new Dictionary<string, int>();
// KeyValuePair<string, int> currentConsensus = new KeyValuePair<string, int>("z", -1);
// foreach (AASeq aaSeq in aaSeqs)
// {
// if (i < aaSeq.Count)
// {
// string residue = aaSeq.SubSeqAA0Pos(i, 1).ToString();
// charToCount[residue] = SpecialFunctions.GetValueOrDefault(charToCount, residue) + 1;
// if (charToCount[residue] > currentConsensus.Value)
// {
// currentConsensus = new KeyValuePair<string, int>(residue, charToCount[residue]);
// }
// }
// }
// consensusSeq.Append(currentConsensus.Key);
//}
AASeq consensusAaSeq = GetAaSeqConsensus(aaSeqs);
NamedSequence consensus = new NamedSequence("consensus", consensusAaSeq.ToString());
return(consensus);
}
new static public AASeq GetInstance(string dnaSeqAsString, MixtureSemantics mixtureSemantics, int offset)
{
DnaSeq dnaSeq = new DnaSeq(dnaSeqAsString, mixtureSemantics, offset);
return(dnaSeq);
}
public void WriteAsTable(List <NamedSequence> sequences, TextWriter writer)
{
CaseIdToAASeq cidToAASeq = CaseIdToAASeq.GetInstance();
bool isDna = sequences[0].IsDna();
foreach (NamedSequence seq in sequences)
{
cidToAASeq.Add(seq.Name,
isDna ?
DnaSeq.GetInstance(seq.Sequence, MixtureSemantics) :
AASeq.GetInstance(seq.Sequence, MixtureSemantics));
}
List <string> header = new List <string>(sequences.Count + 1);
header.Add("Var");
header.AddRange(sequences.Select(seq => seq.Name));
writer.WriteLine(header.StringJoin("\t"));
int maxLen = cidToAASeq.Dictionary.Values.Select(aaSeq => aaSeq.Count).Max();
for (int pos0 = 0; pos0 < maxLen; pos0++)
{
foreach (char aa in cidToAASeq.EveryAminoAcid(pos0))
{
string merAndPos = (pos0 + 1) + "@" + aa;
int?[] values = new int?[sequences.Count];
HashSet <int> nonMissingValues = new HashSet <int>();
for (int pidIdx = 0; pidIdx < sequences.Count; pidIdx++)
{
int? value;
Set <char> observedAAs = cidToAASeq.Dictionary[sequences[pidIdx].Name][pos0];
if (observedAAs.Contains('?') || observedAAs.Count == 0 ||
(observedAAs.Count > 1 && MixtureSemantics == MixtureSemantics.Uncertainty && observedAAs.Contains(aa)))
{
value = null;
}
else if (observedAAs.Contains(aa) && (MixtureSemantics != MixtureSemantics.Pure || observedAAs.Count == 1))
{
value = 1;
}
else
{
value = 0;
}
values[pidIdx] = value;
if (value != null)
{
nonMissingValues.Add((int)value);
}
}
if (nonMissingValues.Count > 1 || (KeepOneValueVariables && nonMissingValues.Count == 1 && nonMissingValues.First() == 1))
{
writer.WriteLine(Helper.CreateTabString(merAndPos, values.Select(v => v.HasValue ? v.ToString() : MissingStatistics.GetInstance().ToString()).StringJoin("\t")));
}
}
}
writer.Flush();
}