Welcome to Auto3D’s documentation!

Note

AIMNet2 clarification: The default model in Auto3D is AIMNet2 since version 2.2.1. If you specify optimizing_engine="AIMNET", it uses AIMNet2. The old AIMNet model has been deprecated since Auto3D 2.2.1.

Auto3D

Introduction

Auto3D is a Python package for generating low-energy conformers from SMILES/SDF. It automates the stereoisomer enumeration and duplicate filtering process, 3D building process, fast geometry optimization and ranking process using ANI and AIMNet neural network atomistic potentials. Auto3D can be imported as a Python library, or be executed from the terminal.

Contents

Getting Started

• Installation
  • Minimum Dependencies Installation
  • Optional Dependencies Installation
• Migrating from v2 to v3
  • Breaking Changes
  • Default Value Changes
  • New Features in v3.0
  • Migration Checklist
  • Getting Help
• Usage
  • Command Line Interface (CLI)
  • Python API
  • Wrapper Functions
  • Parameter Reference
  • Output Files
• CLI Reference
  • Quick Start
  • Global Options
  • Commands
  • Shell Completion
  • Configuration File Format
  • Legacy Mode
  • Exit Codes
  • Environment Variables
  • Examples
• Advanced Usage
  • Multi-GPU Processing
  • Performance Optimization
  • Optimization Parameters
  • Custom Neural Network Potentials
  • Tautomer Enumeration
  • Available Models
  • Troubleshooting
• How-To Guides
  • Quick Start Guide
  • Drug Discovery Workflows
  • Custom Neural Network Potentials
  • High-Performance Computing Guide
  • Troubleshooting Guide
  • Integration with Other Tools

Basic Tutorials

• Auto3D Quick Start
  • CLI Quick Start (Recommended)
  • CLI Commands Reference
  • Python API (for Programmatic Access)
  • 2. Inspect the Results
  • 3. Visualize the Conformers
  • Next Steps
• Tutorial: Getting Optimized Conformers and Energies
  • Quick Start with CLI (Recommended)
  • CLI Reference
  • Python API (for Programmatic Access)
  • Configuration Files
• Performance Tuning Guide
  • CLI Quick Reference (Recommended)
  • Python API Performance Guide
  • 1. Choosing the Right Engine
  • 2. GPU Acceleration
  • 3. Single Model vs Ensemble (AIMNET)
  • 4. TF32 Acceleration (Ampere+ GPUs)
  • 5. Batch Size Optimization
  • 6. Convergence Tuning
  • 7. Reducing Initial Conformers
  • 8. Multi-GPU Processing
  • 9. Memory Profiling
  • Summary: Optimal Settings by Use Case
  • CLI Quick Reference for Performance
• Large-Scale Conformer Generation
  • 1. Memory and Chunking Configuration
  • 2. Optimal Configuration for Large Datasets
  • 3. Monitoring Progress
  • 4. Multi-GPU Processing
  • 5. Processing in Batches
  • 6. Merging Results
  • 7. Resource Monitoring
  • 8. Cleanup
  • Summary: Best Practices for Large-Scale Processing
  • CLI Commands for Large-Scale Processing

Drug Discovery

• Virtual Screening Library Preparation
  • CLI Quick Start (Recommended)
  • Python Workflow
  • 1. Library Preprocessing
  • 2. Lipinski’s Rule of Five Filtering
  • 3. PAINS Filtering
  • 4. 3D Conformer Generation for Docking
  • 5. Post-processing for Docking
  • 6. Conformer Analysis
  • Summary
• Tautomer and Protomer Analysis for Drug Discovery
  • Chemistry Background
  • 1. Common Tautomeric Systems in Drugs
  • 2. Enumerating Tautomers with Auto3D
  • 3. Calculating Tautomer Stability
  • 4. Protonation State Analysis
  • 5. pH-Dependent Conformer Generation
  • 6. Histidine: A Complex Case
  • 7. Integration with Auto3D Tautomer Enumeration
  • 8. Practical Guidelines
  • Summary
• Stereochemistry in Drug Discovery
  • Chemistry Background
  • 1. Famous Stereochemistry Examples in Drugs
  • 2. Identifying Chiral Centers
  • 3. Enumerating Stereoisomers
  • 4. 3D Structure Generation Preserving Stereochemistry
  • 5. Comparing Enantiomer Energies
  • 6. Double Bond Stereochemistry (E/Z)
  • 7. Auto3D Stereoisomer Enumeration
  • 8. Best Practices for Stereochemistry in Drug Discovery
  • Summary
• Docking Integration
  • Workflow Overview
  • 1. Prepare Ligands with Auto3D
  • 2. Convert to PDBQT (AutoDock Vina)
  • 3. Vina Docking Configuration
  • 4. Alternative: MOL2 for GOLD
  • 5. Post-Docking Analysis
  • 6. Complete Workflow Example
  • Summary

Computational Chemistry

• Reaction Thermodynamics with Auto3D
  • Chemistry Background
  • 1. Simple Reaction: Ester Hydrolysis
  • 2. Isomerization Reactions
  • 3. Combustion Reactions
  • 4. Drug Metabolism: Glucuronidation
  • 5. Temperature Dependence of ΔG
  • 6. Reaction Energy Diagrams
  • 7. Best Practices for Reaction Thermodynamics
  • Summary
• Boltzmann Populations and Conformational Analysis
  • Statistical Mechanics Background
  • 1. Basic Boltzmann Analysis
  • 2. Conformational Analysis of a Drug Molecule
  • 3. Boltzmann-Weighted Properties
  • 4. Temperature Effects on Populations
  • 5. Conformational Clustering
  • 6. Free Energy Surface
  • 7. Comparison with Single-Conformer Analysis
  • 8. Practical Applications
  • Summary
• Strain Energy Analysis
  • Chemistry Background
  • 1. Ring Strain: Cycloalkane Series
  • 2. Conformational Strain: Axial vs Equatorial
  • 3. Steric Strain: Gauche Interactions
  • 4. Angle Strain Analysis
  • 5. Drug Molecule Strain
  • 6. Binding Strain
  • 7. Summary
• 3D Molecular Descriptors for ML/QSAR
  • Descriptor Categories
  • 1. Generate 3D Structures
  • 2. 2D Descriptors (Constitutional & Topological)
  • 3. 3D Descriptors (Geometrical)
  • 4. Surface Area Descriptors
  • 5. Energy-Based Descriptors from Auto3D
  • 6. Distance-Based Descriptors
  • 7. Combined Descriptor Table
  • 8. Molecular Fingerprints
  • 9. Descriptor Selection for QSAR
  • 10. Best Practices
  • Summary

Integration Workflows

• Molecular Dynamics Preparation
  • Workflow
  • 1. Generate Optimized Structure
  • 2. Convert to PDB Format
  • 3. GROMACS Topology with ACPYPE
  • 4. GROMACS Solvation and Box Setup
  • 5. Amber Topology with Antechamber
  • 6. OpenMM Python Interface
  • 7. Best Practices
  • Summary
• Quantum Chemistry Refinement
  • Why Use Auto3D for QM?
  • 1. Generate Optimized Structures
  • 2. Gaussian Input Generation
  • 3. ORCA Input Generation
  • 4. Psi4 Input (Python Script)
  • 5. Method/Basis Set Recommendations
  • 6. Batch Input Generation
  • 7. Output Parsing (Example)
  • Summary

Additional Examples

• SPE calculation
• Geometry optimization
• Thermodynamics
  • Generate low-energy conformers with Auto3D
  • Calculate thermodynamic properties with the 3D structures
• Tautomers Tutorial
  • CLI Quick Start (Recommended)
  • Python API
  • 1. Method
  • 2. Example: Get Tautomers Using the Python Library Interface
  • 3. CLI Alternative
• Tautomer Enumeration with Custom NNP
• Using custom NNPs with Auto3D

Reference

• API Reference
  • Core Functions
  • Configuration
  • Model Creation
  • Isomer Generation
  • Tautomer Enumeration
  • Utility Functions
  • Exceptions
• Citation
  • Cite Auto3D
  • Cite AIMNet2
  • Cite ANI-2x
  • Cite ANI-2xt