Getting Started with Chemetrian

Updated 04/20/2026

Welcome to Chemetrian! We are enabling chemists to leverage machine learning in their research and development workflows.

This guide will serve to:
1. Educate on machine learning in chemistry.2. Describe how to use Chemetrian's tools.

Table of Contents

1. Foundations of machine learning in chemistry

  • Molecular Featurization - Learn how to convert molecular structures into numerical representations suitable for ML algorithms
  • Bayesian Optimization - Learn how Bayesian Reaction Optimization can optimize reaction parameter spaces with a minimal number of experiments

2. Chemetrian Tools

  • Descriptor Calculation - Compute quantum chemical and classical cheminformatics descriptors for molecules
  • Reaction Recommender - Get AI-powered reaction condition suggestions with detailed rationale and supporting references
  • Bayesian Reaction Optimization - Optimize large parameter spaces with a minimum number of experiments by leveraging active learning
  • Chemical Space Analysis - Visualize, explore, and filter chemical diversity using dimensionality reduction and clustering techniques
  • Predictive Modeling - Build robust models for molecular property prediction using state-of-the-art algorithms

3. Getting Started

About Chemetrian

Chemetrian is a comprehensive machine learning platform designed specifically for chemical research. Our platform combines cutting-edge ML techniques with domain expertise to accelerate discovery and optimization in chemistry. Whether you're predicting molecular properties, analyzing chemical space, or optimizing reaction conditions, Chemetrian provides the tools and education you need to integrate ML into your research workflow.

Platform Philosophy

We believe that education is fundamental to successful ML adoption in chemistry. Our platform not only provides powerful tools but also comprehensive learning resources that help chemists understand both the theoretical foundations and practical applications of machine learning in their field. This documentation serves as both a user guide and an educational resource for incorporating ML into chemical research.

1. Foundations of machine learning in chemistry

Molecular Featurization

Learn how to convert molecular structures into numerical representations suitable for ML algorithms. This fundamental step is crucial for any machine learning application in chemistry, as it transforms complex molecular information into a format that algorithms can process effectively.

Read more about Molecular Featurization →

Bayesian Optimization

Learn how Bayesian Reaction Optimization can optimize reaction parameter spaces with a minimal number of experiments relative to design of experiments or traditional approaches.

Read more about Bayesian Optimization →

2. Chemetrian Tools

Descriptor Calculation

Upload molecular libraries in SMILES, CDXML, MOL2, or XYZ format and run DFT-based descriptor pipelines. Configure conformer search, geometry optimization, level of theory, solvent, and atom/bond-level features. Download per-molecule and per-atom results for use in ML workflows.

Read more about Descriptor Calculation →

Reaction Recommender

Get AI-powered reaction condition suggestions by entering your reaction type, starting materials, and any constraints. Receive ranked condition sets with detailed rationale and supporting literature references, and download results as a PDF.

Read more about Reaction Recommender →

Bayesian Reaction Optimization

Optimize large parameter spaces with a minimum number of experiments by leveraging active learning.

Read more about Bayesian Reaction Optimization →

Chemical Space Analysis

Visualize, explore, and filter chemical diversity using dimensionality reduction and clustering techniques. This tool helps you understand the relationships between different molecules and identify patterns in your chemical data.

Read more about Chemical Space Analysis →

Predictive Modeling

Build machine learning models to predict molecular properties and chemical outcomes. Upload training data, select features, train models using various algorithms, and make predictions on new molecules before testing them in the lab. Use feature importance analysis to understand the fundamentals driving your chemistry.

Read more about Predictive Modeling →

Getting Help

For additional support, visit our Support Center or contact our team.