Set of small-molecule chemicals found within a biological sample

  • Genome(DNA) > Transcriptome(RNA) > Proteome(Proteins) > Metabolome(Metabolities) > Phenome(Metabolic syndorme, Psychiatric disease) image

Network Analysis

  • (Based on Omics) Genomics, Transcriptomics, Proteomics, Lipidomics, Metabolomics

Scaling Analysis




Open Databases for Drug Development

  • KEGG(Kyoto Encyclopedia of Genes and Genomes)
  • CTD(Comperative toxicogenomics)
  • ChEMBL(Chemical European Molecular Biology Laborotory)




Drug Development Process

  • Target Discovery
    • Expression analysis
    • in-vitro function, in-vivo validation(knockout)
    • Bioinformatics
  • Discovery & Screening
    • Discovery : Structural drug desgin
    • Screening : In-vitro, Ex&In-vivo
  • Lead Optimization
    • Traditional medicinal chemistry
    • Retional drug desgin
  • ADMET
    • BA
    • Systemic exposure
    • Characteristic for ADMET
  • Development
    • Clinic Phase
  • Registration
  • Market

Cost of Target Discovery

Drug > Target(Therapeutic Effect), off-Target(Side Effect)

Drug-Target Network

  • Drug, Target, Disease, Gene
  • Interaction Type
    • Drug-Target, Drug-Drug, Target-Gene, Disease-Disease, Drug-Disease, …

CADD : Computer-Aided Drug Discovery

  • SBDD, Structure-based drug design(Direct drug design) : Molecular docking
  • LBDD, Ligand-based drug design(Indirect drug design) : Quantitative structure-activity relationship (QSAR)




Molecular Binding under Statistical Thermodynamic Perspective

Definition of Free Energy

[REF|5, Calculation of Binding Free Energies]

Gibbs free energy gradient acts as a driving force for a biological system. $$\begin{align*} F &= U-TS = -\frac{1}{\beta}\ln{Q(N,V,T)} \\ G &= H-TS = -\frac{1}{\beta}\ln{Q(N,P,T)} \\ \end{align*}$$

Gibbs free energy

$$\Delta G_{bind} = -k_{B}T\ln{\frac{C^{o}}{8\pi^{2}}\frac{\sigma_{P}\sigma_{L}}{\sigma_{PL}}\frac{Z_{PL}}{Z_{P}Z_{L}}} + P^{o}\Delta V_{PL}$$

  • $k_{B}$ : Boltzman Constant
  • $T, P, V$ : Temperature, Pressure, Volume
  • $C^{o}$ : Standard concentration (1 M)
  • $Z$: Partition functions
  • $\sigma$: Symmetry numbers

Bennet’s Acceptance Ratio (BAR)

From end state equilibrium sampling $$\Delta G_{bind} = \frac{1}{\beta} \frac{\left <f(H_{A}-H_{B}+C)\right >_{B}}{\left <f(H_{B}-H_{A}-C) \right >_{A}} + C$$ $$f(x)=\frac{1}{1+e^{\beta x}},\quad C=\frac{1}{\beta}\ln{\frac{Q_{A}}{Q_{B}}\frac{n_{B}}{n_{A}}}$$

Thermodynamic integration (TI)

Alchemical transition $$\Delta G_{bind} = \int_{0}^{1} \left \langle \frac{\partial H}{\partial \lambda} \right \rangle_{\lambda}\, d\lambda$$

The Quasi-Harmonic Approximation

$$\begin{align*} \Delta G_{bind} & = -k_{B}T \ln \frac{K}{V_{ref}} \\ & = k_{B}T \ln {(8\pi^{2} V_{ref})} - k_{B}T \ln {\int {H(\mathbf{r},\boldsymbol{\Omega})e^{-\beta \omega(\mathbf{r},\boldsymbol{\Omega})}}\, d\mathbf{r}d\boldsymbol{\Omega}} \\ & = k_{B}T \ln {(8\pi^{2} V_{ref}) + \omega_{min}-\frac{k_{B}T}{2} \ln{((2\pi)^{6}\det{C_{\mathbf{r}, \boldsymbol{\Omega}}}})} \\ \end{align*}$$ - $\Delta G_{bind}$ : Absolute binding free energy
- $k_{B}$ : Boltzmann constant
- $T$ : Absolute temperature
- $V_{ref}$ : Reference volume in units consistent
- $\beta=\frac{1}{k_{B}T}$
- $\mathbf{r}, \boldsymbol{\Omega}$ : Relative position, orientation
- $H$ : Ensemble average - $C_{\mathbf{r}, \boldsymbol{\Omega}}$: 6 by 6 fluctuation covariance matrix of the three positional and three orientation coordinates

Free Energy Caluations for Lead Optimization

[REF|Free-energy calculations in structure-based drug design]

$$ L+R \rightleftharpoons C $$ $$\Delta F^{0}_{bind} = - RT \ln{K^{0}_{A}}$$ $$K^{0}_{A}=\frac{1}{K^{0}_{D}}=\frac{k_{on}}{k_{off}}=\frac{[C]}{[L][R]}$$

Soft-core potential[REF|Soft-Core Potentials in Thermodynamic Integration. Comparing One- and Two-Step Transformations]

$$V_{ij} = 4\epsilon_{ij}(1-\lambda)^{t} ([\alpha_{LJ}\lambda^{s} + (r_{ij}/\sigma_{ij})^{n}]^{-12/n} - [\alpha_{LJ}\lambda^{s}+(r_{ij}/\sigma_{ij})^{n}]^{-6/n})$$

IC50

$$IC_{50} = K^{I}_{D} \left ( 1 + \frac{[L_{0}]}{D^{L}_{D}} \right )$$

Structure/Binding Affinity Prediction

Experimental Approach

  • NMR Methods for the Determination of Protein–Ligand Interactions
    • Detection and Verification of Ligand Binding
    • Interaction Site Mapping
    • Interaction Models and Binding Affinity
    • Molecular Recognition
    • Structure of Protein–Ligand Complexes
  • X-ray crystallography
  • cryo-electron microscopy

Theoretical Approach

  • MM/PBSA : the molecule mechanics/Poisson–Boltzmann surface area
  • MM/GBSA : the molecule mechanics/generalized Born surface area
  • LIE : Linear Interaction Methods
  • Fragmentation Methods
    • the fragment molecular orbital (FMO) method
    • the polarized continuum model (PCM)
    • Poisson–Boltzmann (PB) solvation
    • the electrostatically embedded pairwise additive (EE-PA) model
    • the molecular fractionation with conjugate caps (MFCC) method
    • the electrostatically embedded generalized MFCC (EE-GMFCC)
    • the polarizable multipole interaction with supermolecular pairs (PMISP) method image [REF|Ligand-Binding Affinity Estimates Supported by Quantum-Mechanical Methods]
  • The QM-cluster approach
  • Continuum-Solvation Methods

Emprical Approach




Protein-Ligand Interaction

image [REF|Metal–ligand interactions in drug design]

Design Considerations

  • Drug Target Protein Classes
    • GPCRs
    • Ion channels
    • Kinases
    • Proteases
  • Molecular Docking
    • Flexible docking / Rigid docking
    • Scoring Function
      • Force Field
      • Empirical scoring(Solvent Accessible Surface Area value, SASA)
      • Knowledge-Based scoring
    • Search Algorithm
      • Lamarckian Genetic Algorithm
      • Shape Matching
      • Evoluationary Optimization
      • Genetic Algorithm
      • Hybrid
      • Local Optimization
      • Simulated Annealing
      • Swarm-Intelligence Algorithm
  • Molecular Interactions
    • protein–ligand(small molecule)
    • protein–protein
    • protein–nucleic acid
    • protein–carbohydrate
    • protein–lipid
  • Binding Site
  • Water Solvation and Docking

Thermodynamic cycle scheme for the confine-and-release by lock and key model

image [REF|The Confine-and-Release Method: Obtaining Correct Binding Free Energies in the Presence of Protein Conformational Change]

$$\Delta G^{o}_{bind} = \Delta G_{conf} + \Delta G^{o}_{bind,C} + \Delta G_{rel}$$ - $\Delta G^{o}_{bind}$ : the true (standard) binding free energy
- $\Delta G_{conf}$ : the free energy of confining the protein to this smaller region of configuration space in the unbound state
- $\Delta G^{o}_{bind,C}$ : the standard binding free energy of the ligand to the confined protein
- $\Delta G_{rel}$ : the free energy of releasing the protein from conformational confinement in the bound state


G protein-coupled receptor

Protein-Ligand Complex

  • VDW interaction
  • hydrogen bond interaction
  • metal-ligand interaction
  • hydrophobic interaction




Evaluation Metric

  • scoring power, ranking power, docking power, screening power




  • X-Score10, AutoDock Vina8, ChemPLP@GOLD15, GlideScore




Druggability Prediction

Additionals

Big Firm in Pharmaceutical Industry

  • Johnson & Johnson
  • Pfizer
  • Santen
  • Merck
  • Novartis
　

　

　

Reference
  1. List of protein-ligand docking software
  2. AutoDock Vina Installation (Tutorial)
  3. Computational Chemistry
  4. The Relation of State Functions to the Partition Function
  5. Lecture 8: Free energy
  6. Cambridge MedChem Consulting
  7. RCSB PDB Search
  8. FMol
  9. alphafold_pytorch
Related Articles
  1. [Chemistry] Ab initio methods
  2. [Chemistry] Density Functional Theory
  3. [Chemistry] Molecular Mechanics and Dynamics
  4. [Chemistry] QM/MM : Quantum Mechanics/Molecular Mechanics
  5. [Biology] Protein Structure
  6. [Biology] Protein Structure Prediction
  7. [Biology] Enzyme Catalysis
Edit