聚类小分子数据集(基于RDKit的Python脚本)

聚类分子（Clustering molecules）

聚类是一种有价值的化学信息学技术，用于将大型化合物数据集合细分为单个小组相似化合物。其中一个优点是处理非常大的小分子数据集时特别有用。通常用于分析高通量筛选结果、虚拟筛选或对接研究的分析。

基于RDKit的Python脚本用于聚类分子

阅读原文查看完成代码：

#!/usr/bin/python3

def ClusterFps(fps,cutoff=0.2):

from rdkit import DataStructs

from rdkit.ML.Cluster import Butina

# first generate the distance matrix:

dists = []

nfps = len(fps)

for i in range(1,nfps):

sims = DataStructs.BulkTanimotoSimilarity(fps[i],fps[:i])

dists.extend([1-x for x in sims])

# now cluster the data:

cs = Butina.ClusterData(dists,nfps,cutoff,isDistData=True)

return cs

from rdkit import Chem

from rdkit.Chem import AllChem

#generate fingerprints

ms = [x for x in Chem.ForwardSDMolSupplier('ApprovedDrugs.sdf') if x is not None]

fps = [AllChem.GetMorganFingerprintAsBitVect(x,2,1024) for x in ms]

#cluster

clusters=ClusterFps(fps,cutoff=0.4)

# show one of the clusters

print(clusters[20])

#now display structures from one of the clusters

from rdkit.Chem import Draw

from rdkit.Chem.Draw import IPythonConsole

#look at a specific cluster

m1 = ms[1630]

m2 = ms[1010]

m3 = ms[1022]

m4 = ms[1023]

m5 = ms[1034]

m6 = ms[1043]

mols=(m1,m2,m3,m4,m5,m6)

Draw.MolsToGridImage(mols)

参考资料

http://www.rdkit.org/docs/Cookbook.html