public static NEC GetInstanceWithPossibleNulls(int flankingSize, string mer, string protein, int position)
{
string e = protein.Substring(position, mer.Length);
string n;
if (position - flankingSize < 0)
{
//Console.WriteLine("Warning: matching position in protein too close to start to create left flanking region");
n = null;
}
else
{
n = protein.Substring(position - flankingSize, flankingSize);
}
string c;
if (position + mer.Length + flankingSize > protein.Length)
{
//Console.WriteLine("Warning: matching position in protein too close to end to create right flanking region");
c = null;
}
else
{
c = protein.Substring(position + mer.Length, flankingSize);
}
NEC nec = NEC.GetInstance(n, e, c);
return(nec);
}
public static IEnumerable <IHashableFeature> GetNFlankAny1AndNTermEAny2(NEC nec, string nflankAny, string nTermEAny)
{
NFlank nFlankFeature = NFlank.GetInstance();
string nFlankRegion = nFlankFeature.FeatureFunction(nec);
if (-1 == nFlankRegion.IndexOfAny(nflankAny.ToCharArray()))
{
yield break;
}
NTermE nTermEFeature = NTermE.GetInstance();
string nTermERegion = nTermEFeature.FeatureFunction(nec);
if (-1 == nTermERegion.IndexOfAny(nTermEAny.ToCharArray()))
{
yield break;
}
AnyIn anyFeature1 = AnyIn.GetInstance(nflankAny, nFlankFeature);
AnyIn anyFeature2 = AnyIn.GetInstance(nTermEAny, nTermEFeature);
And andFeature = And.GetInstance(anyFeature1, anyFeature2);
Debug.Assert((bool)andFeature.Evaluate(nec)); // real assert - must only generate true features
yield return(andFeature);
}
internal static NEC GetInstance(NEC input, int flankSize)
{
SpecialFunctions.CheckCondition(input.N.Length >= flankSize && input.C.Length >= flankSize, "The input flank size should not be larger than the new flank size");
NEC output = NEC.GetInstance(input.N.Remove(0, input.N.Length - flankSize), input.E, input.C.Substring(0, flankSize));
return(output);
}
public static NEC GetRandomInstance(NEC sample, Dictionary <char, int> aaToCount, ref Random random)
{
string n = RandomAAString(aaToCount, sample.N.Length, ref random);
string e = RandomAAString(aaToCount, sample.E.Length, ref random);
string c = RandomAAString(aaToCount, sample.C.Length, ref random);
return(NEC.GetInstance(n, e, c));
}
private static string GetAminoAcidFromEpitopePosition(NEC nec, int i)
{
char chAminoAcid = nec.E[i];
string aminoAcid = VirusCount.Biology.GetInstance().OneLetterAminoAcidAbbrevTo3Letter[chAminoAcid];
SpecialFunctions.CheckCondition(aminoAcid != null);
return(aminoAcid);
}
public static NEC GetInstance(string peptide, int nLength, int eLength, int cLength)
{
SpecialFunctions.CheckCondition(peptide.Length == nLength + eLength + cLength, "NEC.GetInstance lengths are wrong");
NEC nec = GetInstance(peptide.Substring(0, nLength), peptide.Substring(nLength, eLength), peptide.Substring(nLength + eLength, cLength));
Debug.Assert(nec.N + nec.E + nec.C == peptide); // real assert
return(nec);
}
public static NEC GetConsensusInstanceWithPossibleNullFlanks(string mer, Set <NEC> necSet, ref Random random)
{
DebugCheckMerSameLengthAsEInSet(mer, necSet);
string n = FindConsensusOrNull(necSet, GetN, ref random);
string c = FindConsensusOrNull(necSet, GetC, ref random);
NEC nec = NEC.GetInstance(n, mer, c);
return(nec);
}
static public NEC GetInstance(string n, string e, string c)
{
NEC nec = new NEC();
nec.N = n;
nec.E = e;
nec.C = c;
nec._asString = SpecialFunctions.CreateDelimitedString(",", n, e, c);
return(nec);
}
//Dropping strong typing so that can handle both NEC and Pair<NEC,HLA>
public string FeatureFunction(object entity)
{
NEC nec = entity as NEC;
if (null == nec)
{
nec = ((Pair <NEC, Hla>)entity).First;
}
return(nec.E);
}
public override bool Equals(object obj)
{
NEC other = obj as NEC;
if (other == null)
{
return(false);
}
else
{
return(other._asString.Equals(_asString));
}
}
public static NEC GetInstance(string necString)
{
string[] fieldCollection = necString.Split(',');
SpecialFunctions.CheckCondition(fieldCollection.Length == 3);
for (int iField = 0; iField < fieldCollection.Length; ++iField)
{
if (fieldCollection[iField] == "null")
{
fieldCollection[iField] = null;
}
}
NEC nec = NEC.GetInstance(fieldCollection[0], fieldCollection[1], fieldCollection[2]);
return(nec);
}
public static IEnumerable <IHashableFeature> GetPropertySeqInRegionInstance(int k, Pair <NEC, Hla> necAndHla, IHashableFeature regionFeature)
{
NEC nec = necAndHla.First;
string region = (string)((Feature)regionFeature).Evaluate(necAndHla);
Set <string> aaSeqSet = Set <string> .GetInstance(SpecialFunctions.SubstringEnumeration(region, k));
foreach (string aaSeq in aaSeqSet)
{
foreach (List <string> propertyCombination in KmerProperties.EveryPropertyCombination(aaSeq))
{
string propertySeq = SpecialFunctions.Join(",", propertyCombination);
InProperty feature = InProperty.GetInstance(propertySeq, regionFeature);
Debug.Assert((bool)feature.Evaluate(necAndHla)); // real assert - must only generate true features
yield return(feature);
}
}
}
public string FeatureFunction(NEC nec)
{
string aaSeq = nec.E.Substring(0, 2);
return(aaSeq);
}
public static IEnumerable <IHashableFeature> GetAndNFlankNotNTermEEnumeration(NEC nec)
{
NFlank nFlankFeature = NFlank.GetInstance();
string nFlankRegion = nFlankFeature.FeatureFunction(nec);
Set <string> aaSeqSet = Set <string> .GetInstance(SpecialFunctions.SubstringEnumeration(nFlankRegion, 1));
foreach (string aaSeq in aaSeqSet)
{
Not notFeature = Not.GetInstance(In.GetInstance(aaSeq, NTermE.GetInstance()));
//Only generate the feature when the amino acid is not in NTermE
if ((bool)notFeature.Evaluate(nec))
{
In inFeature = In.GetInstance(aaSeq, nFlankFeature);
And andFeature = And.GetInstance(inFeature, notFeature);
Debug.Assert((bool)andFeature.Evaluate(nec)); // real assert - must only generate true features
yield return(andFeature);
}
}
}
static public string GetN(NEC nec)
{
return(nec.N);
}
public string FeatureFunction(NEC nec)
{
string aaSeq = nec.E.Substring(nec.E.Length - 1);
return(aaSeq);
}
static public string GetE(NEC nec)
{
return(nec.E);
}
private static void AddEiFeatures(bool includeChemicalProperties, bool includeAAFeatures, bool substractChemAACrissCrossFeatures, NEC nec, Pair <NEC, Hla> necAndHla, Set <IHashableFeature> hlaishFeatureSet, Set <IHashableFeature> featureSet)
{
for (int i = 0; i < nec.E.Length; ++i)
{
IHashableFeature featureE = E.GetInstance(i + 1);
string aminoAcid = GetAminoAcidFromEpitopePosition(nec, i);
if (includeAAFeatures)
{
IsAA featureAA = IsAA.GetInstance(aminoAcid, featureE);
featureSet.AddNew(featureAA);
Debug.Assert((bool)featureAA.Evaluate(necAndHla)); // real assert - must only generate true features
foreach (IHashableFeature hlaishFeature in hlaishFeatureSet)
{
if (substractChemAACrissCrossFeatures && hlaishFeature is HasAAProp)
{
continue;
}
And featureHlaishAndAA = And.GetInstance(hlaishFeature, featureAA);
featureSet.AddNew(featureHlaishAndAA);
Debug.Assert((bool)featureHlaishAndAA.Evaluate(necAndHla)); // real assert - must only generate true features
}
}
if (includeChemicalProperties)
{
foreach (string property in VirusCount.KmerProperties.AaToPropList[aminoAcid])
{
HasAAProp featureAAProp = HasAAProp.GetInstance(property, featureE);
featureSet.AddNew(featureAAProp);
Debug.Assert((bool)featureAAProp.Evaluate(necAndHla)); // real assert - must only generate true features
foreach (IHashableFeature hlaishFeature in hlaishFeatureSet)
{
if (substractChemAACrissCrossFeatures && hlaishFeature is IsAA)
{
continue;
}
And featureHlaishAndAAProb = And.GetInstance(hlaishFeature, featureAAProp);
featureSet.AddNew(featureHlaishAndAAProb);
Debug.Assert((bool)featureHlaishAndAAProb.Evaluate(necAndHla)); // real assert - must only generate true features
}
}
}
}
//All of the above with AA replaced by chemical property of AA
}
//static private Dictionary<Pair<NEC, Hla>, bool> CloseHuman = null;
//GeneratorType.Hla | GeneratorType.Position | GeneratorType.Property | GeneratorType.AndHla | GeneratorType.Zero6Supertype | GeneratorType.AndZero6Supertype
private static Set <IHashableFeature> GenerateFeatureSet(
object entity, string supertypeTableSource,
int?flankSizeOrNull,
bool includeFlankNECFeatures,
bool includeChemicalProperties, bool includeAAFeatures,
bool addEiFeatures
)
{
bool includeAndHlaAndSTWithEpitopeAdjFeatures = false;
bool subtractSupertypeFeatures = false;
bool subtractHlaFeatures = false;
bool substractChemAACrissCrossFeatures = false;
SpecialFunctions.CheckCondition(!includeAndHlaAndSTWithEpitopeAdjFeatures || includeFlankNECFeatures);
Pair <NEC, Hla> necAndHlaX = (Pair <NEC, Hla>)entity;
NEC nec = (null == flankSizeOrNull) ? necAndHlaX.First : NEC.GetInstance(necAndHlaX.First, (int)flankSizeOrNull);
Hla hla = necAndHlaX.Second;
Debug.Assert(nec.N.Length == nec.C.Length); // real assert
Pair <NEC, Hla> necAndHla = new Pair <NEC, Hla>(nec, hla);
Set <IHashableFeature> hlaishFeatureSet = new Set <IHashableFeature>();
CreateAndAddHlaFeature(subtractHlaFeatures, hla, necAndHla, ref hlaishFeatureSet);
CreateAndAddFeatureSupertype(supertypeTableSource, subtractSupertypeFeatures, hla, necAndHla, ref hlaishFeatureSet, Assembly.GetExecutingAssembly(), Predictor.ResourceString);
Set <IHashableFeature> featureSet = Set <IHashableFeature> .GetInstance(hlaishFeatureSet);
if (addEiFeatures)
{
AddEiFeatures(includeChemicalProperties, includeAAFeatures, substractChemAACrissCrossFeatures, nec, necAndHla, hlaishFeatureSet, featureSet);
}
if (includeFlankNECFeatures)
{
List <IHashableFeature> aaInNFlankFeatureList = new List <IHashableFeature>(In.GetAASeqInRegionInstance(1, necAndHla, NFlank.GetInstance()));
DebugCheckThatEvaluatesToTrue(necAndHla, aaInNFlankFeatureList);
if (includeAAFeatures)
{
featureSet.AddNewRange(aaInNFlankFeatureList); //AA in N flank
featureSet.AddNewRange(In.GetAASeqInRegionInstance(2, necAndHla, NFlank.GetInstance())); //AA1-AA2 in Nflank
featureSet.AddNewRange(SubSeq.GetInSubSeqEnumeration(NFlank.GetInstance(), false, 1, necAndHla)); //AA@x in N flank (numbering is 5 4 3 2 1)
featureSet.AddNewRange(SubSeq.GetInSubSeqEnumeration(NFlank.GetInstance(), false, 2, necAndHla)); //AA1-AA2@x in Nflank (x is position of AA2, i.e., the smaller number)
featureSet.AddNewRange(In.GetAASeqInRegionInstance(1, necAndHla, CFlank.GetInstance())); //AA in Cflank
featureSet.AddNewRange(In.GetAASeqInRegionInstance(2, necAndHla, CFlank.GetInstance())); //AA1-AA2 in Cflank
featureSet.AddNewRange(SubSeq.GetInSubSeqEnumeration(CFlank.GetInstance(), true, 1, necAndHla)); //AA@x in C flank (numbering is 1 2 3 4 5)
featureSet.AddNewRange(SubSeq.GetInSubSeqEnumeration(CFlank.GetInstance(), true, 2, necAndHla)); //AA1-AA2@x in Cflank (x is position of AA1, i.e., the smaller number)
}
if (includeChemicalProperties)
{
featureSet.AddNewOrOldRange(InProperty.GetPropertySeqInRegionInstance(1, necAndHla, NFlank.GetInstance()));
featureSet.AddNewOrOldRange(SubSeq.GetInPropertySubSeqEnumeration(NFlank.GetInstance(), false, 1, necAndHla));
featureSet.AddNewOrOldRange(InProperty.GetPropertySeqInRegionInstance(1, necAndHla, CFlank.GetInstance()));
featureSet.AddNewOrOldRange(SubSeq.GetInPropertySubSeqEnumeration(CFlank.GetInstance(), true, 1, necAndHla));
featureSet.AddNewOrOldRange(InProperty.GetPropertySeqInRegionInstance(2, necAndHla, NFlank.GetInstance()));
featureSet.AddNewOrOldRange(SubSeq.GetInPropertySubSeqEnumeration(NFlank.GetInstance(), false, 2, necAndHla));
featureSet.AddNewOrOldRange(InProperty.GetPropertySeqInRegionInstance(2, necAndHla, CFlank.GetInstance()));
featureSet.AddNewOrOldRange(SubSeq.GetInPropertySubSeqEnumeration(CFlank.GetInstance(), true, 2, necAndHla));
}
}
if (includeFlankNECFeatures)
{
if (includeAAFeatures)
{
//EV in Epitope
AddFeatureWithOptionalAndHlaAndST(In.GetAASeqInRegionInstance(2, necAndHla, Epitope.GetInstance()), includeAndHlaAndSTWithEpitopeAdjFeatures, hlaishFeatureSet, false, ref featureSet);//AA1-AA2 in Epitope
//RR in Epitope[@1-2]
AddFeatureWithOptionalAndHlaAndST(SubSeq.GetInSubSeqEnumeration(Epitope.GetInstance(), true, 2, necAndHla), includeAndHlaAndSTWithEpitopeAdjFeatures, hlaishFeatureSet, false, ref featureSet);//AA1-AA2@x in Epitope (x is position of AA1, i.e., the smaller number)
}
if (includeChemicalProperties)
{
//polar,cyclic in Epitope
AddFeatureWithOptionalAndHlaAndST(InProperty.GetPropertySeqInRegionInstance(2, necAndHla, Epitope.GetInstance()), includeAndHlaAndSTWithEpitopeAdjFeatures, hlaishFeatureSet, false, ref featureSet);
//polar,large in Epitope[@8-9]
AddFeatureWithOptionalAndHlaAndST(SubSeq.GetInPropertySubSeqEnumeration(Epitope.GetInstance(), true, 2, necAndHla), includeAndHlaAndSTWithEpitopeAdjFeatures, hlaishFeatureSet, false, ref featureSet);
}
//AA1-AA2 in Nflank,Epitope, etc
if (null != flankSizeOrNull && (int)flankSizeOrNull > 0)
{
string epitope = (string)Epitope.GetInstance().Evaluate(entity);
SubSeq lastNAAFeature = SubSeq.GetInstance(1, 1, false, NFlank.GetInstance());
string lastNAA = (string)lastNAAFeature.Evaluate(entity);
In inLastNAA = In.GetInstance(lastNAA, lastNAAFeature);
SubSeq firstEAAFeature = SubSeq.GetInstance(1, 1, true, Epitope.GetInstance());
string firstEAA = (string)firstEAAFeature.Evaluate(entity);
Debug.Assert(firstEAA == epitope.Substring(0, 1));// real assert
In inFirstEAA = In.GetInstance(firstEAA, firstEAAFeature);
SubSeq lastEAAFeature = SubSeq.GetInstance(epitope.Length, epitope.Length, true, Epitope.GetInstance());
string lastEAA = (string)lastEAAFeature.Evaluate(entity);
In inLastEAA = In.GetInstance(lastEAA, lastEAAFeature);
SubSeq firstCAAFeature = SubSeq.GetInstance(1, 1, true, CFlank.GetInstance());
string firstCAA = (string)firstCAAFeature.Evaluate(entity);
In inFirstCAA = In.GetInstance(firstCAA, firstCAAFeature);
if (includeAAFeatures)
{
And andLastNNAAFirstEAA = And.GetInstance(inLastNAA, inFirstEAA);
AddFeatureWithOptionalAndHlaAndST(andLastNNAAFirstEAA, includeAndHlaAndSTWithEpitopeAdjFeatures, hlaishFeatureSet, /*checkThatNew*/ true, ref featureSet);
And andLastEAAFirstCAA = And.GetInstance(inLastEAA, inFirstCAA);
AddFeatureWithOptionalAndHlaAndST(andLastEAAFirstCAA, includeAndHlaAndSTWithEpitopeAdjFeatures, hlaishFeatureSet, /*checkThatNew*/ true, ref featureSet);
}
if (includeChemicalProperties)
{
foreach (string lastNProperty in KmerProperties.AaToPropList[Biology.GetInstance().OneLetterAminoAcidAbbrevTo3Letter[lastNAA[0]]])
{
InProperty inLastNProperty = InProperty.GetInstance(lastNProperty, lastNAAFeature);
foreach (string firstEProperty in KmerProperties.AaToPropList[Biology.GetInstance().OneLetterAminoAcidAbbrevTo3Letter[firstEAA[0]]])
{
InProperty inFirstEProperty = InProperty.GetInstance(firstEProperty, firstEAAFeature); //!!!get this out of the loop?
And andLastNPropertyFirstEProperty = And.GetInstance(inLastNProperty, inFirstEProperty);
AddFeatureWithOptionalAndHlaAndST(andLastNPropertyFirstEProperty, includeAndHlaAndSTWithEpitopeAdjFeatures, hlaishFeatureSet, /*checkThatNew*/ false, ref featureSet);
Debug.Assert((bool)andLastNPropertyFirstEProperty.Evaluate(necAndHla));
}
}
foreach (string lastEProperty in KmerProperties.AaToPropList[Biology.GetInstance().OneLetterAminoAcidAbbrevTo3Letter[lastEAA[0]]])
{
InProperty inlastEProperty = InProperty.GetInstance(lastEProperty, lastEAAFeature);
foreach (string firstCProperty in KmerProperties.AaToPropList[Biology.GetInstance().OneLetterAminoAcidAbbrevTo3Letter[firstCAA[0]]])
{
InProperty infirstCProperty = InProperty.GetInstance(firstCProperty, firstCAAFeature); //!!!get this out of the loop?
And andlastEPropertyfirstCProperty = And.GetInstance(inlastEProperty, infirstCProperty);
AddFeatureWithOptionalAndHlaAndST(andlastEPropertyfirstCProperty, includeAndHlaAndSTWithEpitopeAdjFeatures, hlaishFeatureSet, /*checkThatNew*/ false, ref featureSet);
Debug.Assert((bool)andlastEPropertyfirstCProperty.Evaluate(necAndHla));
}
}
}
}
}
return(featureSet);
}
public static bool TryCreateInstance(string epitopeString, int aa1PosStart, int aa1PosLast, AASeq aaSeqConsensus, out NEC nec)
{
AASeq aaSeqSubConsensus = aaSeqConsensus.SubSeqAA1Pos(aa1PosStart, epitopeString.Length);
Debug.Assert(aaSeqSubConsensus.OriginalAA1Position(0) == aa1PosStart.ToString()); // real assert
string consensusAsString = aaSeqConsensus.ToString();
int flankingSize = 5;
//Create a string builder to file from back to front
StringBuilder sbN = new StringBuilder(new string(' ', flankingSize));
int sbNIndex = flankingSize - 1;
//Look at the characters of the concensus moving to the left
for (int aa1Pos = aa1PosStart - 1;; --aa1Pos)
{
if (aa1Pos < 1)
{
Console.WriteLine("Warning: for epitope {0}, the epitope position is too close to the left of the protein to create a c region of length 5. Skipping epitope", epitopeString);
nec = null;
return(false);
}
char aa = consensusAsString[aa1Pos - 1];
if (aa == '-')
{
continue;
}
sbN[sbNIndex] = aa;
--sbNIndex;
if (sbNIndex < 0)
{
break;
}
}
StringBuilder sbC = new StringBuilder();
//Look at the characters of the concensus moving to the right
for (int aa1Pos = aa1PosLast + 1; ; ++aa1Pos)
{
if (aa1Pos > consensusAsString.Length)
{
Console.WriteLine("Warning: for epitope {0}, the epitope position is too close to the right of the protein to create a n region of length 5. Skipping epitope", epitopeString);
nec = null;
return(false);
}
char aa = consensusAsString[aa1Pos - 1];
if (aa == '-')
{
continue;
}
sbC.Append(aa);
if (sbC.Length == flankingSize)
{
break;
}
}
nec = NEC.GetInstance(sbN.ToString(), epitopeString, sbC.ToString());
return(true);
}
static public string GetC(NEC nec)
{
return(nec.C);
}