This Title All WIREs
How to cite this WIREs title:
WIREs Comput Mol Sci
Impact Factor: 25.113

Computational structure‐based drug design: Predicting target flexibility

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

The role of molecular modeling in drug design has experienced a significant revamp in the last decade. The increase in computational resources and molecular models, along with software developments, is finally introducing a competitive advantage in early phases of drug discovery. Medium and small companies with strong focus on computational chemistry are being created, some of them having introduced important leads in drug design pipelines. An important source for this success is the extraordinary development of faster and more efficient techniques for describing flexibility in three‐dimensional structural molecular modeling. At different levels, from docking techniques to atomistic molecular dynamics, conformational sampling between receptor and drug results in improved predictions, such as screening enrichment, discovery of transient cavities, etc. In this review article we perform an extensive analysis of these modeling techniques, dividing them into high and low throughput, and emphasizing in their application to drug design studies. We finalize the review with a section describing our Monte Carlo method, PELE, recently highlighted as an outstanding advance in an international blind competition and industrial benchmarks. This article is categorized under: Structure and Mechanism > Computational Biochemistry and Biophysics Software > Molecular Modeling
A quick simplified view of what can be done today in molecular modeling. QM dynamics, for example, can only be applied to small systems and for few femtoseconds. On the other site, huge systems and large propagation times require a CG approach
[ Normal View | Magnified View ]
Different protein–ligand interaction mechanisms: lock–key (a), induced‐fit (b), conformational selection (c), and allosteric models (d)
[ Normal View | Magnified View ]
Binding site search and docking simulation for the Src kinase with the PP1 inhibitor (PDB entry 1QCF). (a) Ligand RMSD evolution to the bound crystal along 1,000 MC steps for a 64‐processor standard PELE job. Notice how the red and light blue trajectories find the active site at the end of the simulation (the blue one reaching ~1 Å heavy atom ligand RMSD). The inset shows the interaction energies with respect to the RMSD, where we see the light blue processor reaching the best energies. (b) The kinase structure showing the six different initial positions of the PP1 ligand in the bulk solvent. The active site is highlighted with Leu354 space fill representation. Note that we superimpose here all the six positions, but each trajectory has only one ligand. (c) Ligand RMSD evolution when using the new adaptive PELE for the same system and trajectories. Notice how the adaptive scheme allows reaching the active site in ~50 MC steps (less than 1 hr), improving almost ×20 the efficiency of standard PELE. (d) Interaction energy for the adaptive simulation where we can clearly identify the bound pose as the best one. The color scheme (black to yellow) describes the epochs' evolution in the adaptive procedure
[ Normal View | Magnified View ]
Several docking and MD‐based techniques use the addition of repulsive forces in the active site in order to map its flexibility. In the left illustration we grew some Lennard–Jones particles into the mineral corticoid nuclear hormone receptor active site, aiming at opening it for ligand docking
[ Normal View | Magnified View ]
Different MD contributions for exploring target flexibility in drug design. (a) Target treatment: to generate a receptor conformational ensemble from MD trajectories. (b) Mechanistic studies: to investigate the ligand binding/unbinding pathway and to derive binding free energies. (c) Post refinement: to simulate the stability, induced‐fit and interaction energies of docking poses, in order to improve pose/ligand ranking and for a better structural characterization
[ Normal View | Magnified View ]
(a) Schematic diagram illustrating the idea behind an ensemble docking approach: generation of receptor snapshots, docking, and consensus (or ranking) selection. (b) Screen capture of an Induced Fit Docking (IFD) job setup with the Maestro graphical interface developed by Schrodinger
[ Normal View | Magnified View ]

Browse by Topic

Software > Molecular Modeling
Structure and Mechanism > Computational Biochemistry and Biophysics

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts