Machine Learning Feature Vector
These descriptors form a numerical representation
of the molecule that can be used as input features
for a machine-learning model.
[ — ]
—
From chemistry to machine learning
A molecular structure is first represented by a
SMILES string.
The chemistry engine interprets the SMILES,
constructs the molecular graph, generates
coordinates and calculates molecular descriptors.
The descriptors become numerical features
X for a machine-learning model.
SMILES → Molecular Graph → Descriptors → X → ML Model → Prediction