Table - PMC

Skip to main content

View full-text article in PMC

. 2021 Nov 29;11(12):1783. doi: 10.3390/biom11121783

Algorithm 1 Segmentation of substructures for molecule

G = (V, E)

Input: SMILES strings of compounds

Output: Vocabulary of substructures C

Get molecule object from SMILES
Number the atoms in the compound molecule
Initialize: vocabulary of substructures $C = ⌀$
Construct $V_{1}$ ← the set of bonds $\in E$
Construct $V_{2}$ ← the set of simple rings of G
for each bond $e_{i}$ in $V_{1}$ do
if $e_{i}$ does not belong to any ring then
add $e_{i}$ to the vocabulary of substructures C
end if
end for
for each ring $r_{i}$ in $V_{2}$ do
for each ring $r_{j}$ in $V_{2}$ do
$i n t e r = r_{i} \cap r_{j}$
if the length of $i n t e r$ ≥ 3 then
$t m p$ ← merge $r_{1}, r_{2}$ to one unique ring
$r_{i}$ ← $t m p$
$r_{j}$ ← $t m p$
end if
end for
end for
remove duplicate substructures from $V_{2}$
add each substructure in $V_{2}$ to the vocabulary of substructures C
return vocabulary of substructures C