CSR SALAD v10.1

Cofactor Specificity Reversal - Semi-Automated Library Design

A computational tool for designing site-directed mutagenesis libraries to reverse cofactor specificity in NAD(P)-dependent enzymes.

Repository

https://github.com/fhalab/CSR-SALAD

Overview

CSR SALAD analyzes protein-cofactor binding sites from PDB structures and designs targeted mutagenesis libraries to switch cofactor specificity between NAD and NADP. The tool:

• Identifies residues involved in cofactor binding
• Classifies residues by their interaction geometry (edge, face, bidentate, etc.)
• Generates optimized degenerate codon libraries for specificity reversal
• Suggests additional positions for activity recovery through saturation mutagenesis

Features

• Automated Binding Site Analysis: Identifies first and second shell residues around NAD(P) cofactors
• Geometric Classification: Categorizes residues based on their 3D orientation relative to the adenine ring
• Smart Library Design: Generates minimal libraries within user-defined size constraints
• Recovery Suggestions: Identifies backing residues and hydrogen-bonding partners for activity recovery
• Flexible Options: Exclude/include glycine-rich motifs, diphosphate-binding residues, and peripheral residues

Installation

Option 1: Using Conda (Recommended)

1. Clone this repository:

git clone https://github.com/fhalab/CSR-SALAD.git
cd CSR-SALAD

2. Create the conda environment from the provided file:

conda env create -f environment.yml

3. Activate the environment:

conda activate csr-salad

4. Launch JupyterLab:

jupyter lab

Option 2: Using pip

1. Clone this repository:

git clone https://github.com/fhalab/CSR-SALAD.git
cd CSR-SALAD

2. Ensure you have Python 3.8+ installed
3. Install dependencies:

pip install -r requirements.txt

4. Launch JupyterLab:

jupyter lab

Option 3: Development Installation

For development or if you want to install as a package:

1. Clone the repository:

git clone https://github.com/fhalab/CSR-SALAD.git
cd CSR-SALAD

2. Install in editable mode:

pip install -e .

Requirements

• Python 3.8 or higher
• Biopython >= 1.79
• NumPy >= 1.20.0
• Pandas >= 1.3.0
• JupyterLab >= 3.0.0
• IPython >= 7.0.0

Usage

Quick Start

1. Open CSR_SALAD_v10.1.ipynb in JupyterLab

2. In the first cell, configure your parameters:

   • infile: Path to your PDB file containing the protein-cofactor complex
   • max_size: Maximum library size (typically 0.5× your screening capacity)
   • ex_motif: Exclude glycine-rich motif residues from the library
   • ex_diphos: Exclude diphosphate-contacting residues
   • ex_periph: Exclude peripheral residues
   • verbose: Show detailed analysis log

3. Run all cells (Cell → Run All)

4. Results will appear at the bottom showing:

   • Library design with degenerate codons
   • Site-saturation mutagenesis targets for activity recovery

Example Configuration

infile = './my_enzyme.pdb'
max_size = 400  # Target library size
ex_motif = True  # Exclude glycine-rich motif
ex_diphos = True  # Exclude diphosphate contacts
ex_periph = False  # Include peripheral residues
verbose = False  # Minimal output

Input Requirements

Your PDB file must contain:

• A complete protein structure (or relevant domain)
• At least one bound cofactor molecule with residue name:
  • NAP or NDP for NADP (designing NAD→NADP library)
  • NAD or NAI for NAD (designing NADP→NAD library)

Output

The tool generates two main outputs:

1. Library Design Table

   • Position and residue type
   • Binding classification (Edge, Face, Bidentate, etc.)
   • Degenerate codon for mutagenesis
   • Encoded amino acid diversity

2. Recovery Targets Table

   • Residues prioritized for site-saturation mutagenesis
   • Priority level (High, Medium, Low)
   • Reason for inclusion (backing residue, H-bonding, excluded from library)

Methodology

CSR SALAD uses a geometric analysis approach:

1. Cofactor Detection: Identifies NAD(P) molecules in the structure
2. First Shell Identification: Finds residues within 4.2 Å of the 2'-phosphate moiety
3. Geometric Classification:
   • Transforms residue coordinates into adenine ring-centered coordinate system
   • Classifies based on 3D position (edge, face, bidentate contacts)
   • Identifies glycine-rich motifs
4. Library Generation:
   • Assigns pre-validated degenerate codons based on wild-type residue and geometry
   • Optimizes library size by selective inclusion/exclusion
5. Recovery Analysis:
   • Identifies backing residues behind the adenine ring
   • Finds second-shell charged residues
   • Detects hydrogen-bonding partners

Citation

If you use CSR SALAD in your research, please cite:

CSR SALAD Version 10.1 (2025)
Adapted by Jackson Cahn
California Institute of Technology

Original methodology:

CSR SALAD Version 8 (2015)
Jackson Cahn
California Institute of Technology

Troubleshooting

Common Issues

"Cofactor not found"

• Ensure your PDB file contains NAP/NDP (for NADP) or NAD/NAI (for NAD)
• Check that the cofactor residue names are correctly formatted

"No phosphate atoms found"

• Verify that your cofactor molecule is complete in the PDB file
• Missing atoms may prevent proper analysis

"No residues classified"

• Try expanding the search radius by setting ex_periph = False
• Check that protein residues are near the cofactor binding site

Library too large

• Increase max_size parameter
• Enable exclusion options (ex_motif, ex_diphos, ex_periph)

License

Copyright California Institute of Technology All rights reserved

Author

Jackson Cahn California Institute of Technology

Version History

• v10.1 (October 2025): Updated analysis algorithms and library designs
• v8 (December 2015): Original implementation

Support

For questions, issues, or suggestions, please open an issue on the GitHub repository.