WPI Predict

Machine Learning for Molecular Property Prediction

WPI Predict is a research platform developed at Worcester Polytechnic Institute to advance molecular property prediction through machine learning, computational chemistry, and cheminformatics.

The platform provides researchers with a unified environment for evaluating predictive models, analyzing molecular structures, and supporting computational chemistry workflows through a modern web interface.

Designed as an extensible research platform, WPI Predict enables new machine learning models, datasets, and scientific tools to be integrated as research continues, creating a scalable foundation for future discovery and collaboration.

Getting Started

  1. Open the WPI Predict homepage.
  2. Enter or draw a SMILES chemical structure.
  3. Select the desired prediction model.
  4. Provide any additional model parameters, if required.
  5. Run the prediction.
  6. Review, print, or export your results.

Current Prediction Models

WPI Predict currently provides several machine learning models for estimating experimentally relevant molecular properties. Additional prediction models will become available as ongoing research continues.

  • LogP
  • Boiling Point
  • Melting Point
  • Enthalpy of Vaporization
  • Enthalpy of Fusion
  • Hansen Solubility Parameters
Research Notice

WPI Predict is an academic research platform under active development at Worcester Polytechnic Institute. The platform is designed to support research, education, and collaboration in computational chemistry through machine learning-based molecular property prediction. Prediction results are intended to assist research workflows and should not replace experimental measurement or scientific validation.