Given your ligand and the pocket size, there may not be a simple answer or a straightforward way use existing pharmacophore generating software. You have to decide if you want to find a ligand that is of similar size, where two of them will bind next to each other. Or do you want to find a single ligand that spans the whole binding site region.

An initial place to start is to try to find similar ligand based on simple molecular descriptors (e.g. via ShaEP, Feature trees, MOE’s MACCS). Carefully selected descriptors might allow you to find a few similar compounds that would bind and stack the way the native ligand does.

However, I would recommend starting even more basic, and form some hypotheses that you can computationally test. A better understanding of the system may help you develop a suitable approach for searching for new ligands. By investigating the system, you will develop some chemical intuition concerning it. Some basic questions I would have for you system are

1) Could the ligands adopt different binding poses under physiological conditions?

2) What are the ligand-protein binding strengths of each ligand alone and when together?

3) What is the ligand-ligand interaction energy?

4) How dynamic is the binding pocket?

5) What are the more dynamically stable ligand-protein interactions?

6) How important are all of those sulfur atoms (if that is what the yellow atoms are), and what changes occur if they are replaced with oxygen atoms?

7) What forces are important for binding? How large is the molecule’s dipole moment?

8) How conformationally labile is the molecule?

Insight into many of these questions can be gain through good MD simulations. Such an investigation would be time consuming, but potentially rewarding.

Narrowing down your grid box and re-docking of both the molecules including critical residues might yield a minimum RMSD conformation. After you can model a common pharmacophore for a large library screening.

Hello.

Dock as many inhibitors as possible. It will increase the pool of docking poses, allowing a better pharmacophore detection.

However, I think you are confusing two things;

1) first, for a pharmacophore, you really only need the molecules. Perform a flexible alignment followed by a pharmacophoric consensus on your inhibitors and check which features are common to all molecules;

2) if you want to use the receptor, you are thinking on protein-ligand interaction features (PLIF), comprising all common features for the top-ranked poses of the selected inhibitors.

I would suggest you to do both. First, start with your molecules ONLY to see which features are common and more important. Then, do the PLIF. Hopefully, you will end with similar pharmacophores,thus corroborating your docking poses.

For more pharmacophores, check my publication: Toward a better pharmacophore description of P-glycoprotein modulators, based on macrocyclic diterpenes from Euphorbia species, doi:10.1021/ci200145p - availabe in ResearchGate

Best wishes,

Ricardo

Thank you dear Rajnikant Namdeo for valuable response.

Ricardo Sir,

I do agree with the approach you suggest and so why I am planning to do pharmacophore modelling using Protein-ligand interaction (PLIF) and flexible ligand based along with QSAR. I have 3 different inhibitors and a series of 30 inhibitors whose inhibitory activity has been checked experimentally. I was planning to dock them all against the protein by using the bound conformation of the inhibitor already bound in the crystal structure of the protien. But I found that 2 molecules of this ligand are bound in different conformations in the same binding pocket. My problem is only how to deal with these two molecules of same ligands in the pocket? Shall I just ignore this situation and dock all the inhibitors in the binding pocket or I should treat this considering special?

Thank you and Regards.

Hello.

Thet is a question only you can answer. If those two molecules have high inhibitory activity, you should not ignore them because they can point to a different direction. However, if they are the weakest inhibitors, then perhaps you can ignore them.

Cheers,

Ricardo

Given your ligand and the pocket size, there may not be a simple answer or a straightforward way use existing pharmacophore generating software. You have to decide if you want to find a ligand that is of similar size, where two of them will bind next to each other. Or do you want to find a single ligand that spans the whole binding site region.

An initial place to start is to try to find similar ligand based on simple molecular descriptors (e.g. via ShaEP, Feature trees, MOE’s MACCS). Carefully selected descriptors might allow you to find a few similar compounds that would bind and stack the way the native ligand does.

However, I would recommend starting even more basic, and form some hypotheses that you can computationally test. A better understanding of the system may help you develop a suitable approach for searching for new ligands. By investigating the system, you will develop some chemical intuition concerning it. Some basic questions I would have for you system are

1) Could the ligands adopt different binding poses under physiological conditions?

2) What are the ligand-protein binding strengths of each ligand alone and when together?

3) What is the ligand-ligand interaction energy?

4) How dynamic is the binding pocket?

5) What are the more dynamically stable ligand-protein interactions?

6) How important are all of those sulfur atoms (if that is what the yellow atoms are), and what changes occur if they are replaced with oxygen atoms?

7) What forces are important for binding? How large is the molecule’s dipole moment?

8) How conformationally labile is the molecule?

Insight into many of these questions can be gain through good MD simulations. Such an investigation would be time consuming, but potentially rewarding.

Hi

First, you understand/analyze the two crystal structures of protein-ligand complex.

- If the ligand bound in a conformationally flexible (or even disorder) region of the protein,there is possibility of ligand might have different mode of binding (conformational states of the protein-ligand complex)

- The the binding pose of green colour is better than red one, if the terminal ring of ligand pose (red colour) is highly accessible to solvent (water) which might not be characteristics of good inhibitor

- The image is not really clear, plz check out the sulfur atom (in green, its looking like free SH and in red, its not looking like free SH). Pardon if I am wrong.

@ Parmeswaran. Perfect observation for point 2. For point 3 it is just due to 3D rotation otherwise in red also it is a free thiol.

@Karl Kirschner Thank you very much sir for valuable suggestions and shedding the light on important points necessary to be known. I am working on those now.

Thanks and Regards.

@Parmeswaran Sir, Thank you very much sir.

Sir, the image i have given here is only of one crystal structure in which two molecules of one ligand (say 1a and 1b) are bound in the active site of the protein.

thank you very much for keen observation and suggesting valuable point. It would definitely help me.

Regards.

@Ricardo sir, Karl Kirschner sir, Parmeswaran Sir

Sir, one molecule of the inhibitor is totally burried in the pocket and another one is exposed to the solvent and as the binding pocket is big for occupying the big substrate, that might be the reason for the binding of two molecules of the inhibitor to fill the pocket properly. Can I say, just by observing the protein-ligand complex, that the pocket might not be much flexible to bind with only one molecule of the inhibitor (smaller than substrate) properly? Here I am attaching the image of inhibitor (red and green color) and substrate (cyan color) in one pocket for comparison.

Please comment.

Thanks and Regards.

Nutan.

You could visit the webpage of inteligand (www.inteligand.com) and try the ligandscout tutorials.

Thanks

Dear Nutan, I was wondering if there a chance that you could reveal the ligand's name?