Molecular Docking: Let the Docking Do the Talking

Want to know whether a lead molecule or a ligand of your choice interacts with a particular protein or a receptor? Do you want this information at your fingertips? All it takes is just a few clicks and key presses on the computer and then out comes a computational prediction. This computational process is also known as molecular docking and is easier than wasting a hefty amount of money and time with tedious laboratory work.

What Is Molecular Docking (MD)?

MD is a computational tool you can use to make predictions of possible ligand receptor interactions. The program evaluates all feasible binding pockets of a lead candidate with its target macromolecule. The receptor can be, for example, a protein with a known 3D structure. The individual binding pockets predicted by a ligand/lead combination are known as binding poses. These poses include both the position of the ligand relative to the receptor and the conformational state of the ligand. See the example below in Fig. 1.

Any docking procedure involves the following key tasks:
1. Characterization of the binding site
2. Positioning of the ligand into the binding site (orienting)
3. Evaluating the strength of interaction for a specific ligand-receptor pose (“scoring”)

Two Types of MD

1. Rigid Lock and Key

One model is the classic lock and key model. For this model the protein/receptor is the “lock” and the ligand/lead is the “key”. Both the internal geometry of the receptor and the ligand are kept fixed during docking. One then finds the correct relative orientation of the “key” which will open up the “lock”. In particular, where on the surface of the lock is the key hole and in which direction do you turn the key after it is inserted so that it will open, etc.

2. Flexible or Induced Fit

Induced fit or flexible docking, is a model in which the ligand is kept flexible and the energy needed for different conformations of the ligand fitting into the protein are calculated. Though this method is more time consuming, it can evaluate many different possible conformations which makes it more reliable. During the course of this process, the ligand and the protein adjust their conformation to achieve an overall “best-fit”. The conformational adjustments result in an overall binding pattern that is referred to as an “induced-fit”.

Prerequisites to Employ MD

1. Protein or Receptor Structure Prep

To start with, one needs to obtain a 3D-structure of the protein (Protein Databank Structures or homology models). Then you add the correct number of H-atoms to the 3D protein structure, particularly, the protons required to define the apt ionization and tautomeric states of the amino acids in the protein.

2. Ligand or Lead Molecule Prep

For the ligand, a reasonable 3D structure is needed as a starting point. This is achieved by assigning the correct atom types via understanding the appropriate ionization states, chiralities and tautomeric states of the ligand. This is important because the particular protonation state and tautomeric form of a given ligand undoubtedly influences its ionic and hydrogen bonding abilities.

3. Docking Methodology

The docking methodology to use mainly depends on the type of search algorithm in play and the scoring function. When we talk about the search algorithm, it is a tool which searches all possible confirmations of the ligand. However, there are different types of search algorithms such as Molecular dynamics, Monte Carlo methods, Distance geometry method, and many more. In contrast, the scoring function depicts the binding energy based on how strong the interaction is. Finally, this assigned score decides the best confirmation of the ligand.

What Are Some Applications of MD?

Molecular docking^1-3 is a hot potato for investigators with its versatile applications in diverse avenues of research such as:
Exploring protein-protein interactions,
Computer aided drug discovery (CADD),
High throughput screening,
Unraveling small molecule interactions with large macromolecules, etc.

MD is important because the binding interaction between a small molecule ligand and an enzyme protein may result in the activation or inhibition of an enzyme. If the protein is a receptor, ligand binding may result in agonism or antagonism. Therefore, docking is most commonly in use for the field of drug design, as most drugs are small organic molecules.  Docking also applies to:

• Hit identification – docking combined with a scoring function can be used for instant screening of a huge databases of potential drugs; these in silico experiments can pick out molecules that are likely to bind to the protein target of interest.
• Lead optimization – docking predicts where and in which relative orientation a drug/lead/ligand binds to a target (also known as the binding mode or pose). This information may be help full in designing more potent, active, and selective analogs.
• Bioremediation – MD can also predict pollutants that are enzyme degradable.

Software to Perform MD

My personal favorite software choice, is AutoDock, that heralds from The Scripps Research Institute and the Olson laboratory. Also, this software is an open source resource with a large number of users. There is now a new version, AutoDock-Vina,  that is currently gaining huge citation and usage. A study on the usage of these different programs shows Autodock, Gold, and Glide to be the highest of those in use.

In summary, this gives you a bitesize towards Molecular Docking. If you need any further clarification do post your comments here. Thank you!

References:

1. Morris, Garrett M., and Marguerita Lim-Wilby (2008). Molecular docking. Molecular modeling of proteins. 443, 365-382.
2.Meng, Xuan-Yu, et al. (2011). Molecular docking: a powerful approach for structure-based drug discovery. Current computer-aided drug design. 7.2, 146-157.
3. Chen, Yu-Chian. Beware of docking! (2015) Trends in pharmacological sciences 36.2, 78-95.