ML-Assisted Terpene Synthase Redesign Work Accepted in ACS Catalysis

We are delighted to share that our latest research on engineering sesquiterpene synthases has been accepted for publication in ACS Catalysis. We sincerely thank the editor and referees for their valuable comments and constructive feedback during the review process.

In this work, we developed DeZyme_score, a deep learning-based workflow that integrates a terpenoid-specific binding affinity score with kinetic metrics for enzyme engineering. To further investigate the molecular basis of catalytic improvement, we employed dynamic cross-correlation matrix (DCCM) analysis to uncover long-range interactions involving the water-channel valve residue V314.

By combining DCCM analysis with DeZyme_score, we efficiently screened a designed variant library and explored the sequence–fitness landscape, successfully identifying both active-site and distal mutations that significantly enhanced catalytic activity.

Our study provides a practical framework for the efficient engineering of sesquiterpene synthases and supports the sustainable biosynthesis of valuable sesquiterpene natural products.

For further details, please refer to the original paper: https://pubs.acs.org/doi/10.1021/acscatal.6c01200