MIT researchers release CheMeleon model for predicting molecular properties
Overview
MIT researchers report CheMeleon, an open-source foundation model for predicting molecular properties, in the Sept. 14 issue of the Journal of Chemical Information and Modeling.
The team says CheMeleon, pre-trained on molecular descriptors of one million PubChem molecules, outperformed other foundation and expert-designed models across 58 test cases when few data points are available.
The paper's authors include MIT graduate student Akshat Shirish Zalte, recent graduate Jackson Burns, Research Scientist Charlles Abreu, Professor William Green, and three researchers from BASF. CheMeleon was built with ChemProp, an open-source machine-learning package also developed and maintained by Green's group at MIT. Anyone can download CheMeleon through ChemProp and fine-tune it for their own tasks.
Written by AI from the articles below · updated Nov 6, 2:35 PM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
- MIT News · AIOfficialMIT researchers report CheMeleon model for predicting molecular properties from scant data
AIMIT researchers report CheMeleon, an open-source foundation model pre-trained on molecular descriptors of one million PubChem molecules, in the Sept. 14 Journal of Chemical Information and Modeling. The team says it outperformed other foundation and expert-designed models across 58 test cases when few data points are available. Users can download CheMeleon through the open-source ChemProp package and fine-tune it for their own tasks.
Heat Trend
Not enough continuous observations to show a trend yet.