Aionics Advanced Molecular Property Prediction Models

Single-substance molecular property prediction models trained using state-of-the-art graph neural networks (GNNs).

Run Predictions

To request a prediction, please enter a valid molecular SMILES string with more than one heavy atom. Select at least one of the available properties. You can select multiple properties at once in a single request.

A successful request will return your SMILES, the requested properties, and the standard deviation of the ensemble predictions for each property.

Select properties

Selected Properties

Prediction Results

Sample Input and Output

The input for the model is a single SMILES string with more than one heavy atom.

Example

CCOC(=O)OCC

The model returns a predicted value and a standard deviation estimate for each property of interest.

Example

Flash Point - CCOC(=O)OCC: 23.5 +/- 5.3 C