CMI Project 1.2.11: New in-silico molecular design methods for improved separations

News & Highlights

All News | All Highlights

composite image: Computer-aided design software that predicts selectivity using a machine learning model and then ranks them based on selectivity is now available on GitHub.

Research Highlight

AI algorithm and dataset for metal-binding selectivity predictions

Visual representations of the descriptors included to train LOGKPREDICT: A, atom-to-atom connectivity; B, coordination number; C, rotatable bonds; D, ionic radius; E, strain energy; F, Hancock covalency; G, Hancock electrostatics.

Research Highlight

AI enhances CMI’s computer-aided molecular-design software

Research Highlight

Tools for ligand design now more powerful

Project Team

Principal Researchers

Marilu Perez

marilu@ameslab.gov

Marilu Perez at Ames Laboratory leads the CMI project "New in-silico molecular design methods for improved separations"

This project intends to enable, for the first time, the deliberate design of organic ligands with predetermined metal ion selectivity. To achieve this objective, the project will develop two connected computational methods, involving machine learning and molecular mechanics that predict absolute log K values for the formation of metal-ligand complexes in aqueous solution to an accuracy of +/- 0.3 log K units. Novel log K predicting software will be implemented within existing HostDesigner code to yield the next generation ligand design software capable of identifying novel ligand compositions with optimal affinity and selectivity for a targeted metal ion species.

See more about HostDesigner and other open source software developed by CMI researchers link