Molecular docking apps

[polymath]

Yesterday, I tried to find online ligand-protein docking simulators to see it they could tell the difference between known dopamine reuptake inhibitors (methylphenidate, altropane, amfonelic acid...) and non-DRIs (ibuprofen, cyclizine, citalopram and some random organic compounds). I downloaded the crystallographic structure of the DA transport protein from PDB.

Doing the docking simulation with the DockThor application, it seemed that none of the output parameters (total energy, van der Waals energy and so on) correlated with the difference between the two kinds of compounds. I didn't start specifying any protonated functional groups in the ligand or protein, nor did I give the simulation any clue about how the ligand should approach the protein molecule, it was just a "blind" docking simulation.

Anyone here knows more about this subject? How much do I need to adjust the simulation parameters by hand to get reliable results?

 

[Gaffy]

Let me know if this works out, I'm interested by your work.

 

[polymath]

I also took a look at the playmolecule.org:s KDeep simulator, but preparing the ligand and protein files seems to require several steps... It will probably take some time to learn to use those if it actually requires finding the binding site from the protein 3D structure and predicting the way how the ligand has to approach it.

Edit: I don't have any bioscience training, just physics, mathematics and chemistry (and programming). I know how those docking simulations work, they are just classical mechanics calculations of the ligand moving in the approximate electrostatic field of the protein. The problem here is knowing the list of issues that have to be addressed before I can get useful results from it.

 

[Gaffy]

Where did you find the protein stuctures? For what receptors do you have them?

 

[polymath]

The structures of receptor proteins (for which they are known) are stored on this site: http://www.rcsb.org/ .

 

[Gaffy]

Thanks

 

[Gaffy]

Did it work out?

 

[polymath]

I didn't bother trying to do those docking simulations right now... Maybe later.

Even if a compound is a dopamine reuptake inhibitor, you can't really predict the subjective effects without animal testing. It could be something like pemoline, which just increases your attention span a bit (and can cause liver damage as a side effect),

Anyway, the product of epimyrtine reacting with benzyl grignard reagent is predicted to be more DAT selective than the phenyl version, just like the related compound 4-benzylpiperidine (where the methyl, hydroxyl and the other fused ring are missing).

SwissTargetPrediction

www.swisstargetprediction.ch

Edit: Here's an article about the testing of similar compounds as DAT ligands. However, the chemicals there have only 2-carbon distance between the nitrogen and the aromatic ring.

https://pubs.acs.org/doi/10.1021/jm061354p

 

[Abject]

Yeah affinities do not directly correlate to the subjective experience, drugs affect more than the receptors we study/know
Cool thread and idea tho op

 

[polymath]

In addition to the epimyrtine and lupinine alkaloids I mentioned in the random molecule thread, there is this compound called sedamine, which doesn't seem to have any bioassay data but is a possible serotonin and dopamine reuptake inhibitor according to the Swiss target prediction application.

Sedum acre - Wikipedia

en.wikipedia.org

SwissTargetPrediction

www.swisstargetprediction.ch

https://pubchem.ncbi.nlm.nih.gov/compound/Sedamine

There is some conflicting information about the toxicity of the sedum acre plant that contains this substance. Some say it's a local irritant and possible muscarinic agonist, other sources claim there are no reports of animal poisoning by this plant. Even if it is a dopamine reuptake inhibitor, you can't know whether the natural stereoisomer is the correct one for activity.

Edit: Here's a doctoral thesis about a similar 2-phenethylpiperidine alkaloid, lobelane, and its derivatives which seem to affect dopamine transport in a way that prevents stronger stimulants like methamphetamine from having an effect if taken at the same time:

https://pdfs.semanticscholar.org/eba7/c6647aa7785b28dd9fec5a97b83a93701757.pdf

I also added, in the Wikipedia article about lobeline, a remark about the lack of binding affinity data related to the similar compound sedamine.

 

[S.J.B.]

You might find this article of interest.

 

[polymath]

You might find this article of interest.

Thanks a lot, I'll take a look.