Try Open Babel, http://openbabel.org/docs/dev/Fingerprints/intro.html

Also try ChemMine: http://chemmine.ucr.edu/

Lots of tools, heat mappping, and will export tre files

Try TOXMATCH -http://ihcp.jrc.ec.europa.eu/our_labs/predictive_toxicology/qsar_tools/toxmatch

which contain various similarity calculating methods.

Report the similarity between two structures

http://ctr.wikia.com/wiki/Report_the_similarity_between_two_structures

Particularly, I like this solution:

from rdkit import Chem,DataStructs

mol1 = Chem.MolFromSmiles("CC(C)C=CCCCCC(=O)NCc1ccc(c(c1)OC)O")

mol2 = Chem.MolFromSmiles("COC1=C(C=CC(=C1)C=O)O")

fp1 = Chem.RDKFingerprint(mol1)

fp2 = Chem.RDKFingerprint(mol2)

print "RDK fingerprint: ",DataStructs.TanimotoSimilarity(fp1,fp2)

The output is

RDK fingerprint: 0.471502590674

I did through python RDkit fingerprint following by PCA plot. Finally, clustering into groups.

Naravut Suvannang Hi Naravut, How do you exactly manage to use a PCA and clustering from fingerprints? I am working on something related to that but I am having problems creating those plots.

Nick Wagner Can u give me some info about that problems. By the way, the methods that Waldeyr Silva demonstrated is work and u can adapt to make the heat map.

I think you can find their fingerprint and then calculate their cosine similarity

Hi, I have written a blog on using RDKit generating molecular fingerprint and comparison of molecular structure. It might be helpful to you,

https://medium.com/@prashantkbio/use-machine-learning-to-solve-problems-in-chemistry-dcb9491450c

You can also look in my Jupyter notebook for the python code,

https://github.com/pkumarpl/ML_projects/blob/master/qasr_cheminfo_rdkit/machine-learning-methods-for-drug-activity-prediction-using-rdkit.ipynb

Prashant

hi @all......i have drug smiles with their drugbank ids.Now i need to calculate their chemical structure fingerprints for pairwise similarty based on tanimoto coefficient using rdkit......any help will be highly appreciated

Hi Mir Raas I managed to run rdkit and use for SMILES structures clustering. I can try to help but I need more details on your problem.

@Stanislaw ...Actually the problem is like this.......how to calculate the first drug pairwise similarity based on chemical structure fingerprint corresponding to the 881 chemical substructures defined in PubChem database.

Mir Raas Give a try to this code: https://github.com/labstructbioinf/lbs-tools/blob/master/lbs/ligands/smiles.py

(note that there are other options than TanimotoSimilarity)

Does the Tanimoto or other similarity also implies functional similarity? I am assuming Tanimoto similarity is about structural similarity.

The similarities are based on the molecules' fingerprints (binary representation). There are several methods of computing them. The similarities are simple comparisons of the 1's and 0's. You can use Tanimoto, dice, cousin, Sokal, Russel, Kulczynski, and McConnaughey. RDKit among many other Python libraries can measure such similarities. I have the code if you need it.

Please use the biosimilar tool for the search.

best,