* mark in RECAP, BRICS fragmentation

[hyojin0912]

Hello, thanks for your nice package and management.

I want to get fragment smiles of bulk compounds using RECAP, BRICS module in rdkit.

Following guides (https://buildmedia.readthedocs.org/media/pdf/rdkit/latest/rdkit.pdf, https://www.rdkit.org/docs/source/rdkit.Chem.Recap.html), I got fragments using below codes

SMILES = pd.read_csv('Compound_SMILES.csv')
SMILES = SMILES.drop_duplicates()
SMILES = SMILES.reset_index(drop=True)

# proof and make a list of SMILES
df_smiles = SMILES['SMILES'] # csv input files which contain compounds' SMILES information
c_smiles = []
for ds in df_smiles:
    try:
        cs = Chem.CanonSmiles(ds)
        c_smiles.append(cs)
    except:
        print('Invalid SMILES:', ds)

ms = [Chem.MolFromSmiles(x) for x in c_smiles]

# BRICS 
brics = set()
for m in ms[:10]:
    pieces = BRICS.BRICSDecompose(m)
    brics.update(pieces)

# RECAP
recap = set()
for m in ms[:10]:
    pieces = Recap.RecapDecompose(m).GetAllChildren()
    recap.update(pieces)

Output looks like

brics
{'O=c1[nH]cc(F)c(=O)[nH]1',
 '[1*]C(=O)C(C)C',
 '[1*]C(=O)CCCCCCC(=O)NO',
 '[1*]C([6*])=O',
 '[10*]N1C(=O)c2ccccc2C1=O',

recap
{'*C(=O)C(C)C',
 '*C(=O)CCCCCCC(*)=O',
 '*C(=O)CCCCCCC(=O)NO',

I understand that the first(left) * mark is a dummy atom that notifies cleavage location, so I can delete it to deal with fragment SMILES.
And I thought * mark in the middle stands for any of the elements C, N, O, or S.

>  e.g. brics ->  '[1*]C([6*])=O',
>        recap ->  '*C(=O)CCCCCCC(*)=O',

So I can put C, N, O, or S and regard them as individual fragment SMILES.
Is it right?

Thanks in advance,
Hyojin
Compound_SMILES.csv