分子指纹相似度的可视化

rdkit有一个很炫酷的功能，那就是能可视化显示两个分子的相似性。

以下面两个分子为例：

计算相似度

from rdkit import Chem
from rdkit.Chem import AllChem, DataStructs
from rdkit.Chem.Fraggle import FraggleSim

# define TanimotoSim calculator for convinience.
def calctc(mol1,mol2):
    fp1=AllChem.GetMorganFingerprintAsBitVect(mol1,2)
    fp2=AllChem.GetMorganFingerprintAsBitVect(mol2,2)
    return DataStructs.TanimotoSimilarity(fp1,fp2)

# make molecule from smiles.
mol=Chem.MolFromSmiles("N[C@H](CC(=O)O)C(=O)O")
mol2=Chem.MolFromSmiles("N[C@@H](CO)C(=O)O")

# calc. molecular similarity like ECFP4.
In [26]: calctc(mol,mol2)
Out[26]: 0.3333333333333333
#only N,C difference but low similarity !
 
# calc Fraggle sim.
In [27]:FraggleSim.GetFraggleSimilarity(mol,mol2)
Out[27]: (1.0, '[*]c1ccccc1.[*]c1ccccc1')
# near my feeling.

将相似度映射到分子图像

%matplotlib inline
%pylab inline
from IPython.display import Image
 
from rdkit.Chem import AllChem as Chem
from rdkit.Chem.Draw import IPythonConsole
from rdkit.Chem.Draw import SimilarityMaps
 
smiles1 = 'N[C@H](CC(=O)O)C(=O)O' #ZINC000000895218 (D-Aspartate)
smiles2 = 'N[C@@H](CO)C(=O)O' #ZINC000000895034 (L-Ser)
 
mol1 = Chem.MolFromSmiles(smiles1)
mol2 = Chem.MolFromSmiles(smiles2)
 
SimilarityMaps.GetSimilarityMapForFingerprint(mol2, mol1, SimilarityMaps.GetMorganFingerprint)

结果如下：