Pharmacology Insights into subtype selectivity of opioid agonists by ligand-based and structure-based methods

[AlsoTapered]

Library Genesis

libgen.rocks

Library Genesis

libgen.rocks

Without a shadow of a doubt, the papers in the above hotlinks provide most important data when it comes to the identification and optimization of selective, duel and non-selective opiate agonists as well as high-potency mu ligands.

Fig 1 of the first paper shows the relative spatial position and orientation (RPO) of the key moieties. You will notice that for mu agonists, the paper identifies:

1) 2x hydrogen bond acceptors (HBA)
2) positive ionizable point (PI)
3) ring aromatic group (RA)

This is a limitation as it doesn't identify the SECOND RA (ring aromatic), generally required to achieve very high potency. The second paper DOES identify this second ring.

If you employ the RPO and devise a sufficiently large and varied training set, you will discover that even the most bizarre high-potency ligands fid into these simple rules. However, their is a SIXTH binding site that appears to require pi-bonding and yet does not require a RA. Examples of this are allyl prodine & 14-cinnamyloxycodone although some early synthetic opioids such as alimadol and even the Bentley compounds may have stumbled upon this, The latter may seem an odd example but their are a number of compounds in the class in which the dihydro derivatives are significantly LESS potent.

With the above, it's now possible to state with >98% certainty that any compound with:

1) Biosteric minimum small enough to fid into mu receptor
2) Physical properties (RO5)
3) Ability to cross BBB

WILL be an active opioid.

NB the training-sets are listed from compounds with the highest affinity down to those with the lowest affinity.

 

[AlsoTapered]

I believe that this is a post worthy of being a sticky. Even tianeptine is covered - take a look at ciramadol.

Of course, if someone can find an example that isn't covered, I would be most interested to learn. It's 98%, not 100% which is obviously the goal. Anyone conversant with ChatGPT could probably instruct it to use the papers, the relative positions and then expend the training-sets.

 

[AlsoTapered]

BTW I AM keen for people to suggest mu agonists that do not conform to the training-sets and electronic character's I have provided.

Without such, it's something of a hollow victory AND I only claimed that it covered 98% of ligands. It covers 100% of the ligands I am aware of, but I am by no means concluding that I know of ALL ligands.