Data Science in Chemical Engineering: Applications to Molecular Science.

Area of Science:

• Chemical Engineering
• Data Science
• Artificial Intelligence

Background:

• Chemical engineering is undergoing a significant transformation driven by data science tools.
• Artificial intelligence (AI) applications are poised to impact molecular discovery, design, operations, and manufacturing.
• Molecular data science, focusing on atomic-scale molecular discovery and property optimization, has seen early successes.

Purpose of the Study:

• To summarize key advances in molecular data science for chemical engineers.
• To introduce the concept of a molecular data science life cycle.
• To discuss critical phases within this life cycle.

Main Methods:

• Introduction to the molecular data science life cycle.
• Discussion of data set creation and curation.
• Exploration of molecular representations.
• Overview of data-driven property prediction techniques.
• Examination of de novo molecule generation.
• Consideration of feasibility and synthesizability.

Main Results:

• Highlights advances in applying data science to molecular discovery and property optimization.
• Provides a framework (life cycle) for understanding molecular data science.
• Covers essential stages from data handling to molecule design and validation.

Conclusions:

• Molecular data science is a rapidly advancing subfield with broad implications for chemical engineering.
• Understanding the molecular data science life cycle is crucial for leveraging AI and data science.
• The discussed phases provide a roadmap for future research and application in the field.