public void TestICountFingerprintComparison()
{
var mol1 = TestMoleculeFactory.MakeIndole();
var mol2 = TestMoleculeFactory.MakeIndole();
var fingerprinter = new SignatureFingerprinter();
var fp1 = fingerprinter.GetCountFingerprint(mol1);
var fp2 = fingerprinter.GetCountFingerprint(mol2);
var tanimoto = Tanimoto.Calculate(fp1, fp2);
Assert.AreEqual(1.0, tanimoto, 0.001);
}
public void TestRawTanimotoBetween0and1()
{
var smilesParser = CDK.SmilesParser;
var mol1 = smilesParser.ParseSmiles("Cc1nc(C(=O)NC23CC4CC(CC(C4)C2)C3)c(C)n1C5CCCCC5");
var mol2 = smilesParser.ParseSmiles("CS(=O)(=O)Nc1ccc(Cc2onc(n2)c3ccc(cc3)S(=O)(=O)Nc4ccc(CCNC[C@H](O)c5cccnc5)cc4)cc1");
var fingerprinter = new SignatureFingerprinter(0);
var fp1 = fingerprinter.GetRawFingerprint(mol1);
var fp2 = fingerprinter.GetRawFingerprint(mol2);
var tanimoto = Tanimoto.Calculate(fp1, fp2);
Assert.IsTrue(tanimoto > 0 && tanimoto < 1, $"Tanimoto expected to be between 0 and 1, was:{tanimoto}");
}
public void CompareCountFingerprintAndRawFingerprintTanimoto()
{
var mol1 = TestMoleculeFactory.Make123Triazole();
var mol2 = TestMoleculeFactory.MakeImidazole();
var fingerprinter = new SignatureFingerprinter(1);
var countFp1 = fingerprinter.GetCountFingerprint(mol1);
var countFp2 = fingerprinter.GetCountFingerprint(mol2);
var feat1 = fingerprinter.GetRawFingerprint(mol1);
var feat2 = fingerprinter.GetRawFingerprint(mol2);
var rawTanimoto = Tanimoto.Calculate(feat1, feat2);
var countTanimoto = Tanimoto.Method1(countFp1, countFp2);
Assert.AreEqual(rawTanimoto, countTanimoto, 0.001);
}
public void TestCompaRingBitFingerprintAndCountBehavingAsBit()
{
var mol1 = TestMoleculeFactory.Make123Triazole();
var mol2 = TestMoleculeFactory.MakeImidazole();
var fingerprinter = new SignatureFingerprinter(1);
ICountFingerprint countFp1 = fingerprinter.GetCountFingerprint(mol1);
ICountFingerprint countFp2 = fingerprinter.GetCountFingerprint(mol2);
countFp1.SetBehaveAsBitFingerprint(true);
countFp2.SetBehaveAsBitFingerprint(true);
var bitFp1 = fingerprinter.GetBitFingerprint(mol1);
var bitFp2 = fingerprinter.GetBitFingerprint(mol2);
var bitTanimoto = Tanimoto.Calculate(bitFp1, bitFp2);
var countTanimoto1 = Tanimoto.Method1(countFp1, countFp2);
var countTanimoto2 = Tanimoto.Method2(countFp1, countFp2);
Assert.AreEqual(countTanimoto1, countTanimoto2, 0.001);
Assert.AreEqual(bitTanimoto, countTanimoto1, 0.001);
}
public void TestCountMethod1and2()
{
ICountFingerprint fp1 = new IntArrayCountFingerprint(new Dictionary <string, int>()
{
{ "A", 3 }
});
ICountFingerprint fp2 = new IntArrayCountFingerprint(new Dictionary <string, int>()
{
{ "A", 4 }
});
Assert.AreEqual(0.923, Tanimoto.Method1(fp1, fp2), 0.001);
Assert.AreEqual(0.75, Tanimoto.Method2(fp1, fp2), 0.001);
var mol1 = TestMoleculeFactory.MakeIndole();
var mol2 = TestMoleculeFactory.MakeIndole();
var fingerprinter = new SignatureFingerprinter();
fp1 = fingerprinter.GetCountFingerprint(mol1);
fp2 = fingerprinter.GetCountFingerprint(mol2);
Assert.AreEqual(1.0, Tanimoto.Method1(fp1, fp2), 0.001);
Assert.AreEqual(1.0, Tanimoto.Method2(fp1, fp2), 0.001);
}
/// <summary>
/// Build fingerprint
/// </summary>
/// <param name="mol"></param>
public static BitSetFingerprint BuildBitSetFingerprint(
IAtomContainer mol,
FingerprintType fpType,
int fpSubtype = -1,
int fpLen = -1)
{
// Data for Tanimoto similarity using various fingerprint types for CorpId 123456 query.
// Cart - Standard MDL Oracle Cartridge scores
//
// Similarity Score
// ------------------------------------------------
// Size -> 192 896 1024 1024 128 1024 320
// CorpId Cart MACCS PbChm ECFP4 EXT EState Basic Sbstr
// ------ ---- ---- ---- ---- ---- ---- ---- ----
// 123456 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00
// 123456 0.99 0.98 0.96 0.77 0.95 1.00 0.95 1.00
// 123456 0.99 0.98 0.96 0.77 0.95 1.00 0.94 1.00
// 123456 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00
// 123456 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00
// 123456 0.99 0.91 1.00 0.81 1.00 1.00 1.00 1.00
// 123456 0.98 0.95 1.00 0.74 0.92 1.00 0.93 0.94
// 123456 0.98 1.00 1.00 1.00 1.00 1.00 1.00 1.00
// 123456 0.98 1.00 1.00 1.00 1.00 1.00 1.00 1.00
// 123456 0.98 1.00 0.83 0.76 0.77 0.90 0.76 0.94
// LSH Bin Count - The number of LSH bins (of 25) that match the query bin values
//--------------
// CorpId MAC PbC ECFP EX
// ------ --- --- --- ---
// 123456 25 25 25 25
// 123456 25 20 7 16
// 123456 25 20 9 19
// 123456 25 25 25 25
// 123456 25 25 25 25
// 123456 20 25 9 25
// 123456 21 25 11 17
// 123456 25 25 25 25
// 123456 25 25 25 25
// 123456 25 9 6 11
// Data for Tanimoto similarity using various Circular fingerprint types.
// Using 2 molecules where the 2nd just has an added methyl group.
//
// Measure Score
// -------- -----
// ECFP0 1.00
// ECFP2 .88
// ECFP4 .75
// ECFP6 .64
// FCFP0 1.00
// FCFP2 .92
// FCFP4 .84
// FCFP6 .74
IFingerprinter ifptr = null;
IBitFingerprint ibfp = null;
BitSetFingerprint bfp = null;
IAtomContainer mol2;
string s = "";
DateTime t0 = DateTime.Now;
double getFptrTime = 0, buildFpTime = 0;
if (fpType == FingerprintType.Basic) // size = 1024
{
ifptr = new Fingerprinter();
}
else if (fpType == FingerprintType.Circular) // size variable
{
CircularFingerprinterClass cfpClass = (CircularFingerprinterClass)fpSubtype;
if (cfpClass < CircularFingerprinterClass.ECFP0 || cfpClass > CircularFingerprinterClass.ECFP6)
{
cfpClass = (CircularFingerprinterClass)CircularFingerprintType.DefaultCircularClass; // default class
}
if (fpLen < 0)
{
fpLen = CircularFingerprintType.DefaultCircularLength; // default length
}
ifptr = new CircularFingerprinter(cfpClass, fpLen);
//CircularFingerprinter cfp = (CircularFingerprinter)ifptr;
//ICountFingerprint cntFp = cfp.getCountFingerprint(mol); // debug
//s = CircularFpToString(cfp); // debug
}
else if (fpType == FingerprintType.Extended) // size = 1024
{
ifptr = new ExtendedFingerprinter(); // use DEFAULT_SIZE and DEFAULT_SEARCH_DEPTH
}
else if (fpType == FingerprintType.EState) // size = 128
{
ifptr = new EStateFingerprinter(); // use DEFAULT_SIZE and DEFAULT_SEARCH_DEPTH
}
else if (fpType == FingerprintType.MACCS) // size = 192
{
if (MACCSFp == null)
{
MACCSFp = new MACCSFingerprinter();
}
ifptr = MACCSFp;
}
else if (fpType == FingerprintType.PubChem) // size = 896
{
//IChemObjectBuilder builder = DefaultChemObjectBuilder.getInstance();
ifptr = new PubchemFingerprinter();
}
else if (fpType == FingerprintType.ShortestPath) // size =
{
ifptr = new ShortestPathFingerprinter(); // fails with atom type issue for many structures (e.g. 123456)
}
else if (fpType == FingerprintType.Signature) // size =
{
ifptr = new SignatureFingerprinter(); // can't convert array fingerprint to bitsetfingerprint
}
else if (fpType == FingerprintType.Substructure) // size = 320
{
ifptr = new SubstructureFingerprinter();
}
else
{
throw new Exception("Invalid CdkFingerprintType: " + fpType);
}
getFptrTime = TimeOfDay.Delta(ref t0);
ibfp = ifptr.GetBitFingerprint(mol);
bfp = (BitSetFingerprint)ibfp;
buildFpTime = TimeOfDay.Delta(ref t0);
//long size = bfp.size();
//int card = bfp.Cardinality;
return(bfp);
}
