PDB Structure Fixing

This document provides comprehensive information about HBAT’s PDB structure fixing capabilities, which can automatically enhance protein structures by adding missing atoms, converting residues, and cleaning up structural issues.

Overview

HBAT includes integrated PDB structure fixing capabilities that can significantly improve the quality of structural analysis by:

• Adding missing hydrogen atoms using OpenBabel or PDBFixer

• Adding missing heavy atoms using PDBFixer

• Converting non-standard residues to standard equivalents

• Removing unwanted heterogens while optionally keeping water molecules

• Improving structure quality for more accurate interaction analysis

These capabilities are particularly valuable when working with:

• Crystal structures missing hydrogen atoms

• Low-resolution structures with incomplete side chains

• NMR structures requiring standardization

• Structures containing non-standard amino acid residues

• Structures with unwanted ligands or contaminants

Why PDB Fixing is Important

Most PDB structures from X-ray crystallography lack hydrogen atoms because they are too small to be reliably determined at typical resolutions. Since hydrogen bonds are critical for:

• Protein stability: Secondary and tertiary structure maintenance

• Enzyme catalysis: Active site interactions and mechanism

• Protein-protein interactions: Interface stabilization

• Ligand binding: Drug-target interactions

Accurate hydrogen placement is essential for meaningful interaction analysis.

Supported Methods

HBAT supports two powerful methods for structure enhancement:

OpenBabel

Best for: Basic hydrogen addition with fast processing

Capabilities: - Add missing hydrogen atoms - Handle standard amino acid residues - Fast and lightweight processing - Good for most routine applications

Installation:

conda install -c conda-forge openbabel

Advantages: - Very fast processing - Minimal dependencies - Stable and reliable - Good default hydrogen placement

Limitations: - Cannot add missing heavy atoms - Limited handling of non-standard residues - Basic protonation state handling

PDBFixer

Best for: Comprehensive structure fixing and standardization

Capabilities: - Add missing hydrogen atoms with pH-dependent protonation - Add missing heavy atoms (incomplete side chains) - Convert non-standard residues to standard equivalents - Remove unwanted heterogens - Handle complex structural issues

Installation:

conda install -c conda-forge pdbfixer openmm

Advantages: - Comprehensive fixing capabilities - pH-dependent protonation states - Handles missing heavy atoms - Professional-grade structure preparation - Built-in residue standardization

Limitations: - Larger dependency footprint - Slightly slower processing - More complex for simple tasks

PDB Fixing Parameters

HBAT provides comprehensive control over structure fixing through various parameters:

Core Parameters

Parameter	Default	Type	Description
fix_pdb_enabled	True	Boolean	Enable/disable PDB structure fixing
fix_pdb_method	“openbabel”	String	Method to use: “openbabel” or “pdbfixer”
fix_pdb_add_hydrogens	True	Boolean	Add missing hydrogen atoms
fix_pdb_add_heavy_atoms	False	Boolean	Add missing heavy atoms (PDBFixer only)
fix_pdb_replace_nonstandard	False	Boolean	Convert non-standard residues (PDBFixer only)
fix_pdb_remove_heterogens	False	Boolean	Remove unwanted heterogens (PDBFixer only)
fix_pdb_keep_water	True	Boolean	Keep water molecules when removing heterogens

Advanced Parameters

For PDBFixer method, additional options are available:

Parameter	Default	Description
pH	7.0	pH value for protonation state determination
model_residues	False	Add missing residues to complete chains
keep_ids	True	Preserve original atom numbering

Structure Fixing Logic

When PDB fixing is enabled, HBAT follows this systematic approach:

Processing Pipeline

1. Structure Validation

   • Check input structure integrity

   • Validate atom connectivity

   • Identify missing components

2. Heavy Atom Processing (if enabled)

   • Find missing heavy atoms in residues

   • Add missing side chain atoms

   • Complete incomplete residues

3. Residue Standardization (if enabled)

   • Identify non-standard residues

   • Map to standard equivalents using built-in database

   • Apply custom replacements if specified

4. Heterogen Cleaning (if enabled)

   • Remove unwanted ligands and ions

   • Optionally preserve water molecules

   • Clean up crystal contaminants

5. Hydrogen Addition

   • Determine optimal protonation states

   • Add missing hydrogen atoms

   • Optimize hydrogen positioning

6. Structure Optimization

   • Validate final structure

   • Check for atomic clashes

   • Ensure chemical reasonableness

OpenBabel Logic

OpenBabel uses a straightforward approach:

Input PDB → Parse Structure → Add Hydrogens → Output PDB
                  ↓
             Validate atoms
             Check connectivity
             Apply standard rules

Hydrogen Placement Rules:

• Sp³ carbons: Tetrahedral geometry

• Sp² carbons: Planar geometry

• Nitrogen: Based on hybridization and formal charge

• Oxygen: Lone pair considerations

• Sulfur: Standard coordination patterns

PDBFixer Logic

PDBFixer provides more sophisticated processing:

Input PDB → Find Missing → Add Heavy → Convert → Remove → Add H → Output
            Residues      Atoms      Residues   Hetero   Atoms
               ↓            ↓           ↓         ↓        ↓
            Complete     Side chain   Standard   Clean    pH-based
            chains       completion   residues   structure protonation

Advanced Features:

• pH-dependent protonation: His, Cys, Asp, Glu, Lys, Arg states

• Tautomer handling: His ND1/NE2 protonation

• Metal coordination: Special handling around metal centers

• Disulfide bonds: Proper cysteine pairing

Common Use Cases and Workflows

Basic Hydrogen Addition

Scenario: Crystal structure analysis requiring hydrogen bonds

Recommended Settings:

{
  "fix_pdb_enabled": true,
  "fix_pdb_method": "openbabel",
  "fix_pdb_add_hydrogens": true
}

Example Workflow:

1. Load crystal structure (no hydrogens)

2. Enable PDB fixing with OpenBabel

3. Run analysis with hydrogen bond detection

4. Analyze results with complete hydrogen network

Comprehensive Structure Preparation

Scenario: Drug design requiring pristine protein structure

Recommended Settings:

{
  "fix_pdb_enabled": true,
  "fix_pdb_method": "pdbfixer",
  "fix_pdb_add_hydrogens": true,
  "fix_pdb_add_heavy_atoms": true,
  "fix_pdb_replace_nonstandard": true,
  "fix_pdb_remove_heterogens": true,
  "fix_pdb_keep_water": false
}

Example Workflow:

1. Load raw PDB structure

2. Configure comprehensive fixing

3. Apply all enhancement steps

4. Generate clean structure for analysis

5. Perform interaction analysis on optimized structure

NMR Structure Processing

Scenario: Solution NMR structure requiring standardization

Recommended Settings:

{
  "fix_pdb_enabled": true,
  "fix_pdb_method": "pdbfixer",
  "fix_pdb_add_hydrogens": true,
  "fix_pdb_add_heavy_atoms": false,
  "fix_pdb_replace_nonstandard": true,
  "fix_pdb_remove_heterogens": false,
  "fix_pdb_keep_water": true
}

Example Workflow:

1. Load NMR ensemble (first model)

2. Standardize residue names

3. Add missing hydrogens

4. Preserve native heterogens

5. Analyze with consistent parameters

Membrane Protein Analysis

Scenario: Membrane protein with lipids and detergents

Recommended Settings:

{
  "fix_pdb_enabled": true,
  "fix_pdb_method": "pdbfixer",
  "fix_pdb_add_hydrogens": true,
  "fix_pdb_add_heavy_atoms": true,
  "fix_pdb_replace_nonstandard": false,
  "fix_pdb_remove_heterogens": false,
  "fix_pdb_keep_water": true
}

Rationale: Preserve membrane environment while completing protein structure

Implementation Details

Internal Processing

HBAT’s PDB fixing implementation follows these principles:

Data Flow:

Original PDB → External Tool → Fixed PDB → HBAT Parser → Updated Analysis
     ↓             ↓              ↓            ↓              ↓
Input file    Processing      Enhanced     Complete       Analysis with
              tool            structure    atom set       fixed structure

Direct File Processing:

• File-to-file processing: Direct PDB file enhancement preserving formatting

• Preserved structure: Original PDB formatting and metadata maintained

• Efficient workflow: No intermediate atom-to-PDB conversion needed

• Quality preservation: Professional-grade structure output

Memory Management:

• Direct file processing with minimal memory overhead

• Automatic cleanup of intermediate files

• Error handling with resource protection

• Memory-efficient processing for large structures

Error Handling:

• Tool availability checking

• Parameter validation

• Graceful degradation on failures

• Informative error messages

Quality Control

HBAT implements several quality control measures:

Structure Validation:

• Atom count verification

• Chemical consistency checking

• Geometry reasonableness assessment

• Chain integrity validation

Common Issues Detection:

• Overlapping atoms (clashes)

• Unreasonable bond lengths

• Missing critical atoms

• Inconsistent protonation

Fallback Strategies:

• Alternative method attempts

• Partial processing recovery

• Original structure preservation

• User notification of issues

Performance Considerations

Processing Times (approximate, protein-dependent):

Structure Size	OpenBabel	PDBFixer (Basic)	PDBFixer (Full)
Small (< 100 residues)	< 1 second	1-3 seconds	3-5 seconds
Medium (100-500 residues)	1-3 seconds	3-10 seconds	10-20 seconds
Large (> 500 residues)	3-10 seconds	10-30 seconds	30-60 seconds

Memory Usage:

• Scales roughly linearly with structure size

• PDBFixer requires more memory than OpenBabel

• Temporary file usage for processing

• Automatic cleanup minimizes footprint

Best Practices

Choosing the Right Method

Use OpenBabel when:

• You only need hydrogen atoms

• Processing speed is critical

• Working with standard amino acids

• Simple workflow requirements

Use PDBFixer when:

• Structure has missing heavy atoms

• Non-standard residues are present

• Comprehensive cleanup is needed

• pH-specific protonation is important

Parameter Selection Guidelines

Conservative Approach (minimal changes):

{
  "fix_pdb_enabled": true,
  "fix_pdb_method": "openbabel",
  "fix_pdb_add_hydrogens": true
}

Aggressive Approach (maximum enhancement):

{
  "fix_pdb_enabled": true,
  "fix_pdb_method": "pdbfixer",
  "fix_pdb_add_hydrogens": true,
  "fix_pdb_add_heavy_atoms": true,
  "fix_pdb_replace_nonstandard": true,
  "fix_pdb_remove_heterogens": true,
  "fix_pdb_keep_water": false
}

Balanced Approach (good for most cases):

{
  "fix_pdb_enabled": true,
  "fix_pdb_method": "pdbfixer",
  "fix_pdb_add_hydrogens": true,
  "fix_pdb_add_heavy_atoms": false,
  "fix_pdb_replace_nonstandard": true,
  "fix_pdb_remove_heterogens": false,
  "fix_pdb_keep_water": true
}

Quality Assurance

Before Analysis:

1. Inspect original structure for obvious issues

2. Check resolution and method to set expectations

3. Review heterogen content to plan removal strategy

4. Note any non-standard residues that need handling

After Fixing:

1. Verify atom counts make sense

2. Check for obvious geometry issues

3. Validate critical binding sites are intact

4. Compare before/after for significant changes

Structure Comparison:

• Use structure visualization tools

• Check RMSD of heavy atoms

• Verify preservation of key features

• Examine hydrogen placement quality

Troubleshooting

Common Issues

Tool Not Found:

Error: OpenBabel/PDBFixer is not installed

Solution: Install required dependencies:

# For OpenBabel
conda install -c conda-forge openbabel

# For PDBFixer
conda install -c conda-forge pdbfixer openmm

Processing Failures:

Error: PDBFixer failed: [detailed error]

Common Causes:

• Corrupted input structure

• Unsupported atom types

• Memory limitations

• File permission issues

Solutions:

1. Try alternative method (OpenBabel vs PDBFixer)

2. Simplify fixing parameters

3. Check input file integrity

4. Ensure sufficient disk space

Unexpected Results:

Warning: Atom count changed significantly

Investigation Steps:

1. Check if heterogens were removed unexpectedly

2. Verify non-standard residue conversions

3. Look for added missing atoms

4. Compare before/after structures

Performance Issues

Slow Processing:

• Switch to OpenBabel for speed

• Disable heavy atom addition

• Process smaller structure segments

• Check available memory

Memory Problems:

• Process structures in smaller chunks

• Use OpenBabel instead of PDBFixer

• Ensure adequate swap space

• Close other applications

Integration with Analysis

The PDB fixing functionality integrates seamlessly with HBAT’s analysis pipeline:

Analysis Workflow:

Load PDB → Fix Structure → Parse Fixed PDB → Analyze Interactions → Generate Results
    ↓           ↓                 ↓                  ↓                    ↓
Original    Enhanced          Complete           Accurate             Comprehensive
structure   PDB file          atom set           detection            interaction map
                 ↓
           Fixed PDB Tab
           (GUI Display)

Benefits for Analysis:

• More complete hydrogen bond networks

• Better interaction geometry

• Standardized residue names

• Cleaner structural environment

• More reliable cooperativity detection

• Preserved PDB formatting in output

• Performance metrics and timing information

• GUI integration with Fixed PDB display

Example Analysis Comparison

Without PDB Fixing:

Structure: 1ABC.pdb (no hydrogens)
Total atoms: 1,234
H-bonds detected: 15
Missing interactions due to absent hydrogens

With PDB Fixing:

Structure: 1ABC.pdb (hydrogens added)
Total atoms: 2,108 (+874 hydrogens)
H-bonds detected: 127 (+112 new)
Complete interaction network identified

Future Enhancements

Planned improvements to PDB fixing capabilities:

Enhanced Methods:

• Integration with additional fixing tools

• Custom hydrogen placement algorithms

• Machine learning-based protonation prediction

• Ensemble-aware processing for NMR structures

Performance Optimizations:

• Parallel processing for large structures

• Incremental fixing for structure series

• Caching for repeated processing

• GPU acceleration for compatible operations

Quality Control:

• Automated structure validation metrics

• Before/after comparison reports

• Quality scoring systems

• Integration with structure databases

References and Further Reading

OpenBabel:

• O’Boyle, N.M. et al. “Open Babel: An open chemical toolbox” J. Cheminform. 3, 33 (2011)

• OpenBabel Documentation: http://openbabel.org/docs/

PDBFixer:

• Eastman, P. et al. “OpenMM 4: A Reusable, Extensible, Hardware Independent Library” J. Chem. Theory Comput. 9, 461-469 (2013)

• PDBFixer Documentation: openmm/pdbfixer

Structure Preparation:

• Madhavi Sastry, G. et al. “Protein and ligand preparation: parameters, protocols, and influence on virtual screening enrichments” J. Comput. Aided Mol. Des. 27, 221-234 (2013)

• Shelley, J.C. et al. “A versatile approach for assigning partial charges and valence electron densities in proteins” J. Comput. Chem. 28, 1145-1152 (2007)

---

For questions about PDB fixing functionality or specific use cases, please refer to the HBAT documentation or open an issue on the GitHub repository.