rdkit.Chem.Draw.SimilarityMaps module — The RDKit 2025.03.3 documentation (original) (raw)

rdkit.Chem.Draw.SimilarityMaps.GetAPFingerprint(mol, atomId=-1, fpType='normal', nBits=2048, minLength=1, maxLength=30, nBitsPerEntry=4, **kwargs)¶

Calculates the atom pairs fingerprint with the torsions of atomId removed.

Parameters:

• interest (mol -- the molecule of) –
• for (atomId -- the atom to remove the pairs) –
• fingerprint (fpType -- the type of AP) –
• vector (nBits -- the size of the bit) –
• pair (nBitsPerEntry -- the number of bits available for each) –
• pair –
• pair –

rdkit.Chem.Draw.SimilarityMaps.GetAtomicWeightsForFingerprint(refMol, probeMol, fpFunction, metric=<Boost.Python.function object>)¶

Calculates the atomic weights for the probe molecule based on a fingerprint function and a metric.

Parameters:

• molecule (probeMol -- the probe) –
• molecule –
• function (fpFunction -- the fingerprint) –
• metric (metric -- the similarity) –

Note

If fpFunction needs additional parameters, use a lambda construct

rdkit.Chem.Draw.SimilarityMaps.GetAtomicWeightsForModel(probeMol, fpFunction, predictionFunction)¶

Calculates the atomic weights for the probe molecule based on a fingerprint function and the prediction function of a ML model.

Parameters:

• molecule (probeMol -- the probe) –
• function (fpFunction -- the fingerprint) –
• model (predictionFunction -- the prediction function of the ML) –

rdkit.Chem.Draw.SimilarityMaps.GetMorganFingerprint(mol, atomId=-1, radius=2, fpType='bv', nBits=2048, useFeatures=False, **kwargs)¶

Calculates the Morgan fingerprint with the environments of atomId removed.

Parameters:

• interest (mol -- the molecule of) –
• radius (radius -- the maximum) –
• fingerprint (fpType -- the type of Morgan) – ‘count’ or ‘bv’
• for (atomId -- the atom to remove the environments) –
• vector (nBits -- the size of the bit) –
• false (useFeatures -- if) – ConnectivityMorgan, if true: FeatureMorgan

any additional keyword arguments will be passed to the fingerprinting function.

rdkit.Chem.Draw.SimilarityMaps.GetRDKFingerprint(mol, atomId=-1, fpType='bv', nBits=2048, minPath=1, maxPath=5, nBitsPerHash=2, **kwargs)¶

Calculates the RDKit fingerprint with the paths of atomId removed.

Parameters:

• interest (mol -- the molecule of) –
• for (atomId -- the atom to remove the paths) –
• fingerprint (fpType -- the type of RDKit) – ‘bv’
• vector (nBits -- the size of the bit) –
• length (maxPath -- maximum path) –
• length –
• path (nBitsPerHash -- number of to set per) –

rdkit.Chem.Draw.SimilarityMaps.GetSimilarityMapForFingerprint(refMol, probeMol, fpFunction, draw2d, metric=<Boost.Python.function object>, **kwargs)¶

Generates the similarity map for a given reference and probe molecule, fingerprint function and similarity metric.

Parameters:

• molecule (probeMol -- the probe) –
• molecule –
• function (fpFunction -- the fingerprint) –
• metric. (metric -- the similarity) –
• drawing (kwargs -- additional arguments for) –

rdkit.Chem.Draw.SimilarityMaps.GetSimilarityMapForModel(probeMol, fpFunction, predictionFunction, draw2d, **kwargs)¶

Generates the similarity map for a given ML model and probe molecule, and fingerprint function.

Parameters:

• molecule (probeMol -- the probe) –
• function (fpFunction -- the fingerprint) –
• model (predictionFunction -- the prediction function of the ML) –
• drawing (kwargs -- additional arguments for) –

rdkit.Chem.Draw.SimilarityMaps.GetSimilarityMapFromWeights(mol, weights, draw2d, colorMap=None, scale=-1, size=(250, 250), sigma=None, coordScale=1.5, step=0.01, colors='k', contourLines=10, alpha=0.5, gridResolution=0.1, extraGridPadding=0.5, useFillThreshold=False, fillThreshold=0.01, fillThresholdIsFraction=True, **kwargs)¶

Generates the similarity map for a molecule given the atomic weights.

Parameters:

• interest (mol -- the molecule of) –
• use (weights -- the weights to) –
• object (draw2d -- the RDKit drawing) –
• scheme (colorMap -- the matplotlib color map) –
• map (default is custom PiWG color) –
• scaling (scale -- the) – scale < 0 -> the absolute maximum weight is used as maximum scale scale = double -> this is the maximum scale
• figure (size -- the size of the) –
• Gaussians (sigma -- the sigma for the) –
• coordinates (coordScale -- scaling factor for the) –
• calcAtomGaussian (step -- the step for) –
• lines (alpha -- the alpha blending value for the contour) –
• N (contourLines -- if integer number) – N contour lines are drawn if list(numbers): contour lines at these numbers are drawn
• lines –
• grid (extraGridPadding -- the extra padding of the) –
• grid –
• filled (useFillThreshold -- use a magnitude threshold to determine if a grid box is) –
• boxes (fillThreshold -- the magnitude threshold for filling grid) –
• True (fillThresholdIsFraction -- if) –
• range (the fillThreshold is a fraction of the data) –
• drawing (kwargs -- additional arguments for) –

rdkit.Chem.Draw.SimilarityMaps.GetStandardizedWeights(weights)¶

Normalizes the weights, such that the absolute maximum weight equals 1.0.

Parameters:

weights (weights -- the list with the atomic) –

rdkit.Chem.Draw.SimilarityMaps.GetTTFingerprint(mol, atomId=-1, fpType='normal', nBits=2048, targetSize=4, nBitsPerEntry=4, **kwargs)¶

Calculates the topological torsion fingerprint with the pairs of atomId removed.

Parameters:

• interest (mol -- the molecule of) –
• for (atomId -- the atom to remove the torsions) –
• fingerprint (fpType -- the type of TT) –
• vector (nBits -- the size of the bit) –
• pair (maxLength -- the maxmimum path length for an atom) –
• pair –
• torsion (nBitsPerEntry -- the number of bits available for each) –

any additional keyword arguments will be passed to the fingerprinting function.