7.3 Structural bioinformatics

Structural bioinformatics uses computational methods to analyze and predict the 3D structures of biological molecules. It's crucial for understanding how proteins function and interact, helping researchers design drugs and study diseases at the molecular level.

This field combines biology, chemistry, and computer science to tackle complex problems. From determining protein structures to simulating molecular dynamics, structural bioinformatics offers powerful tools for exploring the microscopic world of biomolecules.

Molecular structure representation

• Molecular structure representation involves describing the spatial arrangement of atoms in a molecule
• Different coordinate systems and hierarchical representations are used to efficiently store and manipulate molecular structures
• Choosing the appropriate representation depends on the specific application and computational requirements

Cartesian coordinate system

Top images from around the web for Cartesian coordinate system

• 

  Frontiers | Molecular dynamics simulation of an entire cell View original

  Is this image relevant?

• 

  File:3D coordinate system.svg - Wikimedia Commons View original

  Is this image relevant?

• 

  Cartesian coordinate system - Wikipedia View original

  Is this image relevant?

• 

  Frontiers | Molecular dynamics simulation of an entire cell View original

  Is this image relevant?

• 

  File:3D coordinate system.svg - Wikimedia Commons View original

  Is this image relevant?

1 of 3

Top images from around the web for Cartesian coordinate system

• 

  Frontiers | Molecular dynamics simulation of an entire cell View original

  Is this image relevant?

• 

  File:3D coordinate system.svg - Wikimedia Commons View original

  Is this image relevant?

• 

  Cartesian coordinate system - Wikipedia View original

  Is this image relevant?

• 

  Frontiers | Molecular dynamics simulation of an entire cell View original

  Is this image relevant?

• 

  File:3D coordinate system.svg - Wikimedia Commons View original

  Is this image relevant?

1 of 3

• Represents each atom's position using x, y, and z coordinates in a three-dimensional space
• Allows for easy calculation of distances and angles between atoms
• Commonly used in molecular dynamics simulations and visualization software
• Example: (1.23, 4.56, 7.89) represents an atom's position in Angstroms

Internal coordinate system

• Describes molecular structure using bond lengths, bond angles, and torsion angles
• Reduces the number of variables needed to represent a molecule compared to Cartesian coordinates
• Useful for describing conformational changes and performing energy minimization
• Example: (1.54 Å, 109.5°, 180°) represents a bond length, bond angle, and torsion angle

Hierarchical vs non-hierarchical representations

• Hierarchical representations organize atoms into a tree-like structure (residues, secondary structure elements)
• Non-hierarchical representations treat all atoms equally without any higher-level grouping
• Hierarchical representations are more efficient for large biomolecules and enable coarse-grained modeling
• Non-hierarchical representations are simpler and more suitable for small molecules or detailed atomic-level simulations

Experimental structure determination methods

• Experimental methods are used to determine the three-dimensional structures of biomolecules at atomic resolution
• Different techniques have their strengths and limitations in terms of sample requirements, resolution, and dynamic information
• Integrating data from multiple methods can provide a more comprehensive understanding of molecular structure and function

X-ray crystallography

• Determines the structure of a molecule by analyzing the diffraction pattern of X-rays scattered by a crystallized sample
• Provides high-resolution structures (typically < 2.5 Å) but requires well-ordered crystals
• Suitable for proteins, nucleic acids, and small molecules
• Example: (Lysozyme) was the first enzyme structure solved by X-ray crystallography

Nuclear magnetic resonance (NMR) spectroscopy

• Determines the structure of a molecule by analyzing the magnetic properties of atomic nuclei in a strong magnetic field
• Provides information on the structure, dynamics, and interactions of molecules in solution
• Limited to relatively small proteins (< 50 kDa) due to spectral complexity
• Example: (HIV-1 protease) structure was determined by NMR, revealing its dimeric nature

Cryo-electron microscopy (Cryo-EM)

• Determines the structure of a molecule by analyzing the electron density maps of flash-frozen samples
• Enables structure determination of large macromolecular complexes and membrane proteins
• Rapidly advancing technique with improving resolution (< 3 Å) and reduced sample requirements
• Example: (SARS-CoV-2 spike protein) structure was quickly solved by Cryo-EM, aiding in vaccine development

Protein structure prediction

• Protein structure prediction involves computationally determining the three-dimensional structure of a protein from its amino acid sequence
• Different approaches are used depending on the availability of template structures and the complexity of the target protein
• Predicted models are evaluated using various quality assessment methods to estimate their reliability

Template-based modeling

• Predicts the structure of a target protein by using homologous proteins with known structures as templates
• Relies on the principle that evolutionarily related proteins often share similar structures
• Includes methods such as homology modeling and fold recognition
• Example: (MODELLER) is a widely used software for template-based modeling

Ab initio modeling

• Predicts the structure of a protein without using any template information
• Samples the conformational space to find the most energetically favorable structure
• Computationally intensive and limited to small proteins (< 150 amino acids)
• Example: (Rosetta) is a popular software suite for ab initio modeling

Hybrid approaches

• Combines template-based and ab initio modeling techniques to improve prediction accuracy
• Uses template information when available and ab initio modeling for regions without suitable templates
• Leverages the strengths of both approaches to handle a wider range of target proteins
• Example: (I-TASSER) is a successful hybrid approach that combines threading and ab initio modeling

Model quality assessment

• Evaluates the quality and reliability of predicted protein models
• Uses various scoring functions based on physical principles, statistical potentials, and machine learning
• Helps in selecting the best models and estimating their accuracy
• Example: (ProQ3D) is a deep learning-based method for assessing the local and global quality of protein models

Structural alignment and comparison

• Structural alignment involves finding the best superposition of two or more protein structures
• Comparison methods quantify the similarity between protein structures and identify conserved regions
• Structure-based function prediction relies on the principle that proteins with similar structures often have similar functions

Pairwise structure alignment

• Aligns two protein structures to minimize the distances between corresponding atoms
• Commonly used algorithms include (DALI), (TM-align), and (CE)
• Provides a measure of structural similarity (RMSD, TM-score) and identifies equivalent residues
• Example: Aligning (hemoglobin) and (myoglobin) reveals their structural similarity despite low sequence identity

Multiple structure alignment

• Simultaneously aligns three or more protein structures to identify conserved regions
• Useful for studying evolutionary relationships and identifying core structural motifs
• Methods include (MUSTANG), (Matt), and (MultiProt)
• Example: Aligning (serine proteases) from different organisms highlights their conserved catalytic triad

Structural similarity measures

• Quantifies the similarity between protein structures using various metrics
• Common measures include (RMSD), (TM-score), (GDT-TS), and (Z-score)
• Helps in assessing the quality of structural alignments and comparing different protein folds
• Example: (TM-score) ranges from 0 to 1, with values > 0.5 indicating a similar fold

Structure-based function prediction

• Predicts the function of a protein based on its structural similarity to proteins with known functions
• Uses structure-function databases (CATH, SCOP) and annotation tools (ProFunc, COFACTOR)
• Complements sequence-based function prediction methods, especially for distantly related proteins
• Example: Predicting the (enzymatic activity) of a novel protein based on its structural similarity to a characterized enzyme

Protein-ligand interactions

• Protein-ligand interactions play a crucial role in many biological processes, such as (enzyme catalysis) and (signal transduction)
• Understanding these interactions is essential for drug discovery and design
• Computational methods are used to predict binding sites, dock ligands, and evaluate binding affinities

Binding site prediction

• Identifies potential ligand binding sites on a protein surface
• Uses geometric, energetic, and evolutionary information to locate cavities and pockets
• Methods include (CASTp), (FPocket), and (SiteMap)
• Example: Predicting the (active site) of an enzyme based on the presence of conserved residues and a deep cavity

Docking algorithms

• Predicts the binding pose of a ligand within a protein's binding site
• Samples the conformational space of the ligand and evaluates the complementarity with the protein
• Popular docking software include (AutoDock), (GOLD), and (Glide)
• Example: Docking a (small molecule inhibitor) into the active site of a (kinase) to predict its binding mode

Scoring functions

• Evaluates the strength of protein-ligand interactions and estimates the binding affinity
• Uses force field-based, empirical, or knowledge-based potentials to calculate interaction energies
• Helps in ranking docked poses and prioritizing compounds for experimental testing
• Example: Using a (consensus scoring) approach to improve the prediction of binding affinities

Virtual screening applications

• Screens large libraries of compounds to identify potential ligands for a target protein
• Combines docking and scoring to rank compounds based on their predicted binding affinity
• Enables the prioritization of compounds for experimental validation, saving time and resources
• Example: Identifying (novel inhibitors) for a (drug target) by virtually screening a library of (FDA-approved drugs)

Molecular dynamics simulations

• Molecular dynamics (MD) simulations predict the time-dependent behavior of molecular systems
• Simulates the motion of atoms and molecules based on Newton's laws of motion and a force field
• Provides insights into the dynamics, conformational changes, and interactions of biomolecules

Force fields and parameterization

• Force fields define the potential energy functions and parameters used in MD simulations
• Common biomolecular force fields include (AMBER), (CHARMM), and (GROMOS)
• Parameterization involves deriving force field parameters from experimental data or quantum mechanical calculations
• Example: Using the (AMBER ff14SB) force field to simulate the dynamics of a (protein-ligand complex)

Simulation protocols and analysis

• MD simulation protocols define the initial conditions, boundary conditions, and simulation parameters
• Typical steps include (system preparation), (energy minimization), (equilibration), and (production run)
• Analysis of MD trajectories provides information on (structural fluctuations), (conformational changes), and (interactions)
• Example: Analyzing the (root-mean-square deviation) (RMSD) of a protein backbone to assess its stability during a simulation

Conformational sampling and free energy calculations

• MD simulations can be used to sample the conformational space of a molecular system
• Enhanced sampling methods (replica exchange, umbrella sampling) improve the exploration of rare events
• Free energy calculations (MM-PBSA, thermodynamic integration) estimate the relative stability of different states
• Example: Using (umbrella sampling) to calculate the (free energy landscape) of a protein folding process

Structural databases and resources

• Structural databases store and organize experimentally determined and computationally predicted molecular structures
• Different databases focus on specific types of molecules (proteins, nucleic acids, small molecules) or structural features
• These resources provide valuable data for comparative analysis, method development, and structure-based drug design

Protein Data Bank (PDB)

• The PDB is the primary repository for experimentally determined 3D structures of biological macromolecules
• Contains (X-ray crystallography), (NMR), and (Cryo-EM) structures of proteins, nucleic acids, and complexes
• Provides a standardized format (PDB format) for representing molecular structures
• Example: Downloading the (crystal structure) of a (target protein) from the PDB for virtual screening

Classification databases (SCOP, CATH)

• SCOP (Structural Classification of Proteins) and CATH (Class, Architecture, Topology, Homology) are hierarchical classification databases
• Organize protein structures into (families), (superfamilies), (folds), and (classes) based on their structural and evolutionary relationships
• Useful for studying protein structure-function relationships and identifying distant homologs
• Example: Using SCOP to identify proteins with similar (folds) but different (functions)

Sequence-structure databases (SWISSPROT, UniProt)

• SWISSPROT and UniProt are comprehensive databases that provide protein sequence and functional information
• Integrate data from various sources, including (structural databases), (literature), and (experimental annotations)
• Useful for mapping sequence-structure-function relationships and identifying homologous proteins
• Example: Retrieving the (UniProt entry) of a protein to access its (sequence), (structure), and (functional annotations)

Structure visualization and analysis tools

• Visualization and analysis tools are essential for exploring, manipulating, and interpreting molecular structures
• Different software packages offer a range of features and user interfaces tailored for specific applications
• Scripting and programming interfaces enable the automation of tasks and the development of custom analysis pipelines

Molecular graphics software

• Molecular graphics software allows the interactive visualization and manipulation of molecular structures
• Popular programs include (PyMOL), (Chimera), and (VMD)
• Enables tasks such as (rendering), (coloring), (labeling), and (measuring distances and angles)
• Example: Using (PyMOL) to visualize the (surface electrostatic potential) of a protein and its ligand binding site

Structural analysis and validation tools

• Structural analysis tools provide quantitative measures of molecular structure properties
• Validation tools assess the quality and consistency of experimental and predicted structures
• Examples include (MolProbity) for geometry validation, (ProCheck) for stereochemical quality, and (WHATCHECK) for atomic contacts
• Example: Using (MolProbity) to evaluate the (Ramachandran plot) and (rotamer outliers) of a protein structure

Scripting and programming interfaces

• Many molecular graphics and analysis tools provide scripting and programming interfaces
• Allow the automation of repetitive tasks and the development of custom analysis workflows
• Common scripting languages include (Python) and (Tcl)
• Example: Writing a (Python script) to automate the (structural alignment) and (RMSD calculation) of multiple protein structures