I studied on a series of pyrimidine derivatives. I had biological study results which shows that their chloro derivative is most active. But the most active compound not have highest binding energy (Docking Score; Autodock Vina). So is there any relationship between activity and binding energy??
Mr. Haritwal,
Yes, there is.
DGBind = RT.ln(IC50) + const.
You should pay attention to the topic: Thermodynamics.
Yes Martin i'm agree with you, but kindly suggest me which factors can effect the binding scores?
More information is needed in order to answer your question. For example, what assay were you using to assess biological activity? What is the presumed protein target for the biological activity? Which protein target were you using for the docking study?
In general, docking is better at predicting the spatial orientation of a ligand in a receptor binding site than it is at predicting the free energy of binding and the associated binding constant.
Note also that the free energy of binding and the "binding constant" (Ki) are reciprocally related, so that the more negative the free energy of binding, the more favorable the binding, whereas the lower the value of Ki, the more favorable is the binding.
There is no relationship between them as docking is binding ability, and all strong binders are not biologically active conformer.
There is no exact correlation between docking score and inhibitory activity.
For more explanation, you can check the link.
Good Luck
https://www.researchgate.net/post/What_would_be_the_reasons_of_the_unmatching_between_docking_scores_and_IC50_values_with_references_please?_tpcectx=qa_overview_following&_trid=H3Cz0h4Xw9M3GpoyLHauDMcI_
Dear Amit
There in no one to one correspondence between docking score and the biological activity. This topic has been discussed several times, please go through the link:
https://www.researchgate.net/post/What_would_be_the_reasons_of_the_unmatching_between_docking_scores_and_IC50_values_with_references_please
Mr. Klavan,
Let us see available nominalisations and connotations of the old 'free Gibbs energy' term known from 1860s.
For example, looking at the text of Ref. 1, it provides the following notions about 'scoring functions':
''Brief review of scoring functions........
....Over the years, different scoring functions have been developed that exhibit different accuracies and computational efficiencies. In this section, we will briefly review the scoring functions in literature developed for proteinâ-ligand interactions in molecular docking...
...Force field scoring function...
...Despite its lucid physical meaning, a major challenge in the force field scoring functions is how to treat the solvent in ligand binding...
...Empirical scoring function..
...A second kind of scoring functions are empirical scoring...
functions, which estimate the binding affinity of a complex on the basis of a set of weighted energy terms....
DGi represents different energy terms such as vdW energy, electrostatics, hydrogen bond, desolvation, entropy, hydrophobicity, etc....'' [Ref. 1]
The DGi represents, thereby, the well-known 'free Gibbs energy'. It reflects the binding energy ("binding affinity", above), because of we are talking about "...vdW, electrostatics, hydrogen bond....", but interactions, treated in energy terms, furthermore, quantitatively.
Here, there is a question to what extent does this, in fact, specific notion ('scoring functions') tell anything new as methodology? At first there may seem to be a new quantitative treatment of the ligand-protein interactions, but the connotations shown above (Ref. 1) appear to suppose that the answer is 'no'. Ref. 1 shows that the notion 'scoring functions' means, in fact, the well-known thermodynamics' methodologies.
Ref. 1. Scoring functions and their evaluation methods for proteinâ-ligand docking: recent advances and future directions, S. Huang, S. Grinter, X. Zou, Phys.Chem.Chem.Phys., 2010,12, 12899-12908.
The protein can be conveniently considered as a solvent. The protein-ligand binding can be viewed as a "solvent-substitution" event. The protein-ligand binding free energy quantifies the ligand's relative preference for the two environments (protein in solution versus solution) at a given temperature and as an ensemble average. The ensembles can be generated computationally using molecular dynamics simulations.
---
"Molecular docking free energy" is just a fancy name with little relationship to reality; it is sometimes numerically correct, but it is always conceptually wrong.
Indeed, ensembles are must if one has to mimic biological environment.
Mr. Klavana,
If your opinion is that the free Gibbs energy is wrong, you should know that the term "Molecular Dynamics Simulations" as written in your posting provides exactly the same "wrong" energy concept. Why?
I shall treat 'protein-ligand molecular interactions' in terms of 'solute - solvent interactions' as you would like. I shall do so by assuming that a protein (P) is a "solvent", while a ligand (L) is a "solute" molecule.
The so-called "solvation" process (or dissolving) between L and P is expressed by:
Dmu*s = DGs*
Where mu*s is the so-called 'chemical potential', while DGs* is the free Gibbs energy term. It accounts for the coupling "work" or the energy needed to "dissolve" a L into a solvent P.
This equation shows why Gibbs free energy cannot be described as a correction. It is the energy value which is computed.
DG*s is given by:
DG*s = -kTln{exp-beta.BE},
Where BE denotes the already mentioned 'binding energy' and beta = (kT)-1.
From a 'molecular level mechanistics' point of view a such as reaction of "dissolving" of a solute into a solvent is associated with different types of intermolecular interactions and corresponding molecular level processes. Or a series of micro-states determins the macroscopic thermodynamic (again) parameters.
For the sake of brevity, I shall, here, only look at a common equation, which is among the most fundamental tenet, treating solute-solvent interactions:
DG*s = GS + GS/Sol + SumiGiS/Sol + Sumi,jGi,jS/Sol+....
Here, the DGs* term is expressed by a sum, including all those possible interaction processes, taking into consideration as well as processes associated with many more known types of inramolecular perturbations of the structures and intra-molecular interaction networks.
The first two terms account for the 'work' (meaning energy) needed of a solute (L) to create the so-called "cavity" into the solvent environment (P). But the process of creation of such as cavity is associated with conformational/intra-molecular changes of the interacting species. The next terms express energies of different (specific and non-specific) solute-solvent interactions (H-bonding, p-stacking, etc...).
Given that, the main goal of a molecular modeling, in general, or docking in particular, is to obtain information about the energetics of the interacting ensembles. This goal is pursued using the methods of quantum chemistry (MDs are only part of these methods).
Depending on the philosophy and the corresponding accuracy of these computational methodologies, there are obtained DG*s quantities with different accuracies. And because of it is not obvious how to explain the 'philosophy' of a 'quantum chemical method', here, there could be said that under 'philosophy', rather, we have reflected in mind theoretical concepts (or mental viewing) of treatment of all these different types of conformational and interaction changes/processes, including, treatment of common environmental factors in a macroscopic terms (For example, computations of P-L ensembles at different T, P, ionic strengths and s.o.).
The thermodynamic phenomenology has universal character. Or the thermodynamics laws are valid to the Universe as whole. Like the Newton's universal law of gravitation, for example.
As you might see, I hope, from my short representation of P-L interactions in terms of 'solute-solvent' interactions, that the essence of this problematic is a quantum chemical treatment of molecular level interactions/processes. Or these are just molecules like all other molecular systems.
I should also note that RG is a just aforum, not University. Therefore there is unable to give participants an overview of all quantum chemical methodologies, even very particularly talking about MD ones.
This is an area of professional qualification in the 'Chemistry', underlining
explicitly the following sub-areas of the Chemistry: 'Theoretical chemistry', 'physical chemistry' and 'chemical physics'.
Simply put: No, but (some cases) Yes!
Depends on the type of bioactivity, ligand and receptor (including their structural, geometrical, electronic, and other physico-chemical properties paraphernalia including the intrinsic and extrinsic factors playing part, yes case by case or type by type!) specifics have to be detailed out from the literature. Don't expect every time for it to be following the prediction. Bioactivity is a play of several factors, try to grab most of them through your SAR/QSAR and play with your prediction and bioactivity observance! Good Luck.
- Badges
- Science topic