Python library for combining smiles 'skeletons' with functional groups & substituents in a combinatorial fashion.
smilescombine provides molecular building blocks (in the form of SMILES strings)
that can be used in combination with Supramolecular structure-building software such
as stk. Alternatively, the bulding blocks can be used to generate
structured molecular libraries in their own right.
Using smilescombine begins by initialising a class Combiner, which accepts a
SMILES string of the molecular skeleton we will combine with functional groups
and a list of those functional groups (substituents, represented by SMILES).
substituents = ['(N(C)C)', '(N)', '(OC)', '(O)', '(CC)', '(C=O)', '(C(=O)OC)']
skeleton = Combiner('c1ccccc1', substituents, nmax=4, nconnect=0, auto_placement=True)
skeleton.combine_substituents()We can then use the combine_substituents() method of Combiner to generate our
molecular library for a this skeleton. With the above arguments, combine_substituents()
will allow a maximum of 4 substitutions and place them automatically on all
accessible aromatic carbon atoms within the skeleton.
Now, the attribute combinations of the instance of the Combiner class skeleton
is a list of SMILES strings containing all possible positional and compositonal
combinations of our molecular skeleton and the substituent SMILES supplied by
substituents.
If automatic placement of substituents is accectable, we can define an ordinary
SMILES string. If we wish to specify where substitutions may take place, we can
specify auto_placement=False within Combiner and supply a custom skeleton
SMILES string:
skeleton = Combiner('c1c(Br)cc(Br)cc1', substituents, nmax=4, nconnect=0, auto_placement=True)Where allowed substitution sites are indicated by '(Br)' within the string.
smilescombine relies on rdkit, which can be installed via conda (recommended)
$ conda install -c rdkit rdkit
To install, simply do pip install --user smilescombine to get the latest version from PyPI.