oepandas 2.1.1

Native handling of OpenEye objects in Pandas

OEPandas

Deep integration of OpenEye objects into Pandas DataFrames with native support for molecules and design units.

---

🚀 Quick Start

pip install oepandas

import oepandas as oepd
from openeye import oechem

# Load molecule data from various formats
df = oepd.read_sdf("molecules.sdf")
df = oepd.read_oeb("molecules.oeb.gz")  
df = oepd.read_molecule_csv("data.csv", molecule_columns="SMILES")

# Use pandas normally with molecules
df["num_oxygens"] = df.Molecule.apply(lambda mol: oechem.OECount(mol, oechem.OEIsOxygen()))

✨ Features

• Native OpenEye Integration: Store OEGraphMol and OEDesignUnit objects directly in pandas DataFrames
• Multiple File Formats: Read SDF, OEB, CSV, SMI, OEDB, and OEDU files seamlessly
• Pandas Extensions: Rich accessor methods for molecular operations (.to_smiles(), .depict(), .get_mols(), etc.)
• Type Safety: Full type hints and PyCharm IDE support
• Performance: Optimized for large molecular datasets

---

📖 Table of Contents

• Installation
• Basic Usage
  • Reading Molecular Data
  • Working with Molecules
  • Design Units
• Advanced Features
  • Custom Accessors
• Examples
• API Reference
• Development

---

🔧 Installation

Requirements

• Python 3.10+
• pandas 2.1.0+
• numpy
• OpenEye Toolkits 2023.1.0+
• more-itertools

Note on OpenEye Toolkits License: OpenEye Toolkits requires a commercial license. However, free licenses are available for academic and non-profit institutions. Visit OpenEye Scientific or contact OpenEye directly to request an academic license.

Install from PyPI

pip install oepandas

Development Installation

git clone https://github.com/scott-arne/oepandas.git
cd oepandas
pip install -e ".[dev]"

---

📚 Basic Usage

Reading Molecular Data

OEPandas provides readers for all major chemical file formats supported by the OpenEye Toolkits, including their proprietary formats and record files:

import oepandas as oepd

# SDF files - molecules with properties
df = oepd.read_sdf("molecules.sdf")
print(df.head())
#                                   Molecule    Title      MolWt
# 0  <oechem.OEGraphMol; proxy...>   Aspirin   180.157
# 1  <oechem.OEGraphMol; proxy...> Ibuprofen   206.281

# CSV files with SMILES
df = oepd.read_molecule_csv("data.csv", molecule_columns="SMILES")

# OEB files (binary format)
df = oepd.read_oeb("molecules.oeb.gz")

# Design unit files
df = oepd.read_oedu("complexes.oedu")

# OERecord databases
df = oepd.read_oedb("records.oedb")

Working with Molecules

Once loaded, use pandas normally with rich molecular extensions:

from openeye import oechem

# Standard pandas operations work
filtered_df = df[df.MolWt > 200]

# Rich molecular accessors
smiles = df.Molecule.to_smiles()
images = df.Molecule.depict(width=300, height=200)

# Apply OpenEye functions
df["oxygen_count"] = df.Molecule.apply(lambda mol: oechem.OECount(mol, oechem.OEIsOxygen()))
df["has_ring"] = df.Molecule.apply(lambda mol: oechem.OEDetermineRingMembership(mol) > 0)

# Convert to different formats
df["canonical_smiles"] = df.Molecule.to_smiles(flavor=oechem.OESMILESFlag_Canonical)

Design Units

Work with protein-ligand complexes:

# Read design unit file
df = oepd.read_oedu("protein_ligand_complexes.oedu")

# Extract components
df["Ligand"] = df.Design_Unit.get_ligands()
df["Protein"] = df.Design_Unit.get_proteins()

# Analyze components
df["ligand_mw"] = df.Ligand.apply(oechem.OECalculateMolecularWeight)
df["protein_residues"] = df.Protein.apply(lambda mol: oechem.OECount(mol, oechem.OEIsResidue()))

---

🔥 Advanced Features

Custom Accessors

OEPandas registers many useful pandas accessors automatically:

# Molecular property accessors
df.Molecule.copy_molecules()          # Deep copy molecules
df.Molecule.to_smiles()              # Generate SMILES strings
df.Molecule.depict()                 # Generate 2D depictions
df.Molecule.as_molecule()            # Convert to different formats

# Design unit accessors  
df.Design_Unit.get_ligands()         # Extract ligand molecules
df.Design_Unit.get_proteins()        # Extract protein molecules
df.Design_Unit.copy_design_units()   # Deep copy design units

# DataFrame-level accessors
df.oechem.write_sdf("output.sdf")    # Write to SDF file
df.oechem.write_oeb("output.oeb")    # Write to OEB file

---

📘 Examples

Comprehensive Jupyter notebooks are available in the examples/ directory:

• 01_getting_started.ipynb - Basic usage, molecular calculations, and data manipulation
• 02_advanced_features.ipynb - File I/O, design units, performance optimization, and ML integration

---

📋 API Reference

File Readers

• read_sdf(filename, **kwargs) - Read SDF files
• read_oeb(filename, **kwargs) - Read OEB files
• read_oedu(filename, **kwargs) - Read OEDU files
• read_molecule_csv(filename, molecule_columns, **kwargs) - Read CSV with molecules
• read_smi(filename, **kwargs) - Read SMILES files
• read_oedb(filename, **kwargs) - Read OEDB files

Core Classes

• MoleculeArray / MoleculeDtype - Pandas extension for molecules
• DesignUnitArray / DesignUnitDtype - Pandas extension for design units
• DisplayArray / DisplayDtype - Pandas extension for molecular displays

---

🛠 Development

Running Tests

invoke test
# or
pytest

Building Package

invoke build
# or  
python -m build

---

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

---

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.

---

👤 Author

Scott Arne Johnson

• Email: [e-mail]

---

🔗 Related Projects

• OpenEye Toolkits - The underlying cheminformatics toolkit
• Pandas - Data analysis library that OEPandas extends

---