DeCAF is a method for describing molecules' pharmacophoric properties and a fast and efective tool for comparing and combining ligands.
DeCAF is written as a Python module and can be easily combined with OpenBabel, RDKit and other chemoinformatic tools.
#!python
#use RDKit
from rdkit.Chem import MolFromSmiles
from decaf.toolkits import rd
from decaf.utils import similarity, combine_pharmacophores, draw
#create models
mol1 = MolFromSmiles("c1cc(cc(c1)N)c1cc(cc(c1O)c1[nH]c2ccc(cc2n1)C(=N)N)Cl")
phar1 = rd.phar_from_mol(mol1)
mol2 = MolFromSmiles("c1cc(cc(c1)N(=O)=O)c1cc(cc(c1O)c1cc2cc(ccc2[nH]1)C(=N)N)CC(=O)O")
phar2 = rd.phar_from_mol(mol2)
#compare and combine models
print similarity(phar1, phar2)
phar = combine_pharmacophores(phar1, phar2)
#draw pharmacophore
draw(phar)
There are also three demo scripts, showing prossible applications of DeCAF:
- compare_demo.py - compare two sets of ligands to refrence set
- filter_demo.py - screen database for molecules similar to given pharmacophore model
- model_demo.py - create pharmacophore models from set of molecules
- numpy (for basic functionalities)
- Pybel (OpenBabel) and/or RDKit (for creating models from molecules)
- matplotlib (for pharmacophore drawing)
- scipy (for spring layout)
Automatic documentation for DeCAF can be build with Sphinx (http://sphinx-doc.org/)
cd docs
make html
make latexpdf
After installing OpenBabel or RDKit, you can install DeCAF with setuptools:
python setup.py install
...or use pip:
pip install decaf
You can also install DeCAF and all its dependencies with conda:
#add conda channels for DeCAF, OpenBabel and RDKit
conda config --add channels marta-sd
conda config --add channels mwojcikowski
conda config --add rdkit
#install DeCAF with all dependencies
conda install decaf
Please contact us with comments, suggestions, questions or bug reports:
- Marta M. Stepniewska-Dziubinska ( martasd[at]ibb.waw.pl )
If you use DeCAF in your research, please cite:
Stepniewska-Dziubinska, M. M., Zielenkiewicz, P., & Siedlecki, P. (2017). DeCAF—Discrimination, Comparison, Alignment Tool for 2D PHarmacophores. Molecules, 22(7), 1128; doi:10.3390/molecules22071128