🧬Bioinformatics Unit 10 – Structural bioinformatics

Introduction to Structural Bioinformatics

• Focuses on the analysis and prediction of the 3D structures of biological macromolecules (proteins, DNA, RNA)
• Combines principles from biology, chemistry, physics, and computer science to understand the relationship between sequence, structure, and function
• Plays a crucial role in drug discovery, protein engineering, and understanding disease mechanisms
• Utilizes experimental techniques (X-ray crystallography, NMR spectroscopy, cryo-electron microscopy) to determine structures
• Develops computational methods to analyze, compare, and predict structures when experimental data is unavailable or limited
• Enables the study of large-scale structural data, such as protein-protein interactions and molecular dynamics simulations
• Contributes to the development of personalized medicine by identifying drug targets and designing targeted therapies

Protein Structure Basics

• Proteins are linear polymers composed of amino acids linked by peptide bonds
• Primary structure refers to the amino acid sequence of a protein
• Secondary structure includes local conformations (α-helices, β-sheets) stabilized by hydrogen bonds
  • α-helices are right-handed spiral conformations with 3.6 amino acids per turn
  • β-sheets are extended conformations with strands connected by hydrogen bonds
• Tertiary structure is the overall 3D shape of a single polypeptide chain
  • Stabilized by interactions between side chains (hydrophobic, electrostatic, hydrogen bonds, disulfide bridges)
• Quaternary structure involves the arrangement of multiple polypeptide chains into a multi-subunit complex
• Protein folding is the process by which a polypeptide chain acquires its native 3D structure
  • Driven by the minimization of free energy and the hydrophobic effect
• Misfolding or aggregation of proteins can lead to various diseases (Alzheimer's, Parkinson's, prion diseases)

Computational Methods for Structure Analysis

• Sequence alignment techniques (pairwise, multiple) identify conserved regions and evolutionary relationships
• Structural alignment methods compare and superimpose 3D structures to identify common folds and motifs
• Molecular visualization tools (PyMOL, Chimera) enable interactive exploration and analysis of structures
• Molecular dynamics simulations study the motion and conformational changes of proteins over time
  • Based on Newton's laws of motion and force fields that describe atomic interactions
• Normal mode analysis identifies low-frequency collective motions that are functionally relevant
• Protein-ligand docking predicts the binding pose and affinity of small molecules to protein targets
• Homology modeling constructs 3D models of proteins based on the structures of related homologs
• Machine learning approaches (deep learning, graph neural networks) are increasingly used for structure prediction and analysis

Sequence-Structure Relationships

• Anfinsen's dogma states that the amino acid sequence determines the native structure of a protein
• Evolutionary conservation of residues often indicates structural or functional importance
• Mutations can affect protein stability, folding, and function
  • Single nucleotide polymorphisms (SNPs) are common genetic variations
  • Missense mutations lead to amino acid substitutions
  • Nonsense mutations introduce premature stop codons
• Sequence motifs are short, conserved patterns that often correspond to functional sites (active sites, binding sites)
• Protein domains are independently folding units that can be combined to form multi-domain proteins
• Intrinsically disordered regions lack stable 3D structures but can adopt specific conformations upon binding
• Sequence-based methods (PSI-BLAST, HHpred) can identify remote homologs and predict structural features

Structure Prediction Techniques

• Ab initio methods predict 3D structures from sequence alone, without relying on known structures
  • Based on physicochemical principles and energy minimization
  • Computationally intensive and limited to small proteins
• Comparative modeling (homology modeling) predicts structures based on sequence similarity to known structures
  • Templates are identified using sequence alignment (BLAST, PSI-BLAST)
  • Models are built by aligning the target sequence to the template structure
• Fold recognition (threading) methods align sequences to known folds based on compatibility scores
• Protein structure prediction servers (Robetta, I-TASSER) integrate multiple approaches and provide automated predictions
• AlphaFold and RoseTTAFold are deep learning-based methods that have achieved high accuracy in recent CASP competitions
• Model quality assessment programs (MolProbity, ProCheck) evaluate the stereochemical quality and plausibility of predicted structures
• Experimental validation (X-ray crystallography, NMR) is essential to confirm predicted structures

Molecular Docking and Drug Design

• Molecular docking predicts the binding pose and affinity of small molecules (ligands) to protein targets (receptors)
• Docking algorithms sample the conformational space and evaluate the complementarity of ligand-receptor interactions
  • Search algorithms include systematic, stochastic, and simulation-based methods
  • Scoring functions estimate the binding free energy based on physicochemical properties
• Virtual screening filters large libraries of compounds to identify potential hits for a given target
• Structure-based drug design optimizes lead compounds based on their interactions with the target structure
  • Iterative process of design, synthesis, and testing to improve potency and selectivity
• Pharmacophore modeling identifies the essential features of ligands that are required for binding and activity
• ADME/Tox properties (absorption, distribution, metabolism, excretion, toxicity) are considered in drug optimization
• Successful examples include HIV protease inhibitors, kinase inhibitors (imatinib), and influenza antivirals (oseltamivir)

Structural Databases and Tools

• Protein Data Bank (PDB) is the primary repository for experimentally determined 3D structures
  • Contains over 180,000 structures of proteins, nucleic acids, and complexes
  • Structures are annotated with metadata (resolution, method, ligands, mutations)
• PDBsum provides a graphical overview and analysis of PDB entries
• UniProt is a comprehensive database of protein sequences and functional annotations
• Pfam is a database of protein families and domains based on sequence alignments and hidden Markov models
• SCOP and CATH are hierarchical classifications of protein structures based on evolutionary and structural relationships
• RCSB PDB, PDBe, and PDBj are worldwide data centers that provide access to PDB data and tools
• Open-source software libraries (BioPython, BioJava, BioPerl) facilitate the development of custom analysis tools

Applications in Research and Industry

• Structural bioinformatics contributes to the understanding of protein function, evolution, and disease mechanisms
• Enables the identification of drug targets and the design of new therapeutics
  • Structure-guided optimization of lead compounds
  • Prediction of off-target effects and toxicity
• Facilitates the engineering of proteins with enhanced stability, specificity, or novel functions
  • Rational design of enzymes for biocatalysis and industrial applications
  • Development of antibodies and other protein-based therapeutics
• Supports the interpretation of genetic variants and their impact on protein structure and function
  • Identification of disease-causing mutations and potential targets for personalized medicine
• Integrates with other omics data (genomics, transcriptomics, proteomics) for systems-level understanding of biological processes
• Collaborations between academia and industry drive the development of new technologies and applications
  • Public-private partnerships (Structural Genomics Consortium) for large-scale structure determination
  • Spin-off companies commercialize tools and services for drug discovery and protein engineering