Finding Pathways in Reaction Networks Guided by Energy Barriers Using Integer Linear Programing.

Area of Science:

• Computational Chemistry
• Chemical Reaction Engineering
• Systems Chemistry

Background:

• Analyzing synthesis pathways in chemical reaction networks is crucial for understanding molecular transformations.
• Existing methods often lack kinetic information or struggle with large, complex networks.

Purpose of the Study:

• To develop a computational methodology for searching and ranking synthesis pathways in chemical reaction networks.
• To enable kinetically informed pathway investigations, even for networks lacking prior kinetic data.

Main Methods:

• Utilized integer linear programming for pathway searching.
• Modeled reaction networks using directed hypergraphs.
• Developed a physically-based objective function to maximize pathway probability.
• Created an automated pipeline for estimating reaction energy barriers.

Main Results:

• Successfully applied the methodology to a reaction network generated from 2-hydroxyethanenitrile, water, and ammonia.
• Identified synthesis pathways for glycine and 2-hydroxyethanoic acid.
• Demonstrated the ability to investigate large networks computationally, even those without kinetic annotations.

Conclusions:

• The developed methodology provides a flexible and computationally efficient approach for kinetically informed pathway analysis.
• This method is applicable to large reaction networks, including those generated by generative models.
• Facilitates the discovery of synthesis routes in complex chemical systems.