Explore the molecular dataset used in this application
This dataset contains 643 molecules clustered into 4 groups based on their electronic and structural properties. The clustering was performed using K-means algorithm with 4 clusters after dimensionality reduction using Principal Component Analysis (PCA).
The molecular data was obtained from Density Functional Theory(DFT), Molecular Dynamics(MD) calculation and deep-learning model prediction. The dielectric constants represent the solvent environment used in the calculations.
| Cluster | Number of molecules | Percentage |
|---|---|---|
| Cluster 1 | 270 | 42.0% |
| Cluster 2 | 48 | 7.4% |
| Cluster 3 | 39 | 6.1% |
| Cluster 4 | 286 | 44.5% |
| Property | Min | Max | Mean | Std Dev |
|---|---|---|---|---|
| \(E_{\textrm{S}}-E_{\textrm{A}}\) (eV) | -0.07 | 2.06 | 0.92 | 0.22 |
| LUMO energy (eV) | -0.75 | 8.61 | -4.26 | 4.03 |
| HOMO energy (eV) | -9.58 | -7.35 | -7.69 | 0.25 |
| Dielectric constant | 1.01 | 58.85 | 5.22 | 6.09 |
Molecular size and shape
Number of fluorine atoms
Solvent polarity
Relative binding ability
Reduction stability
Oxidation stability