Misc How do some people predict how a molecule will feel by looking at it

[polarthedog]

I see some posts where a guys like “well it has the nitrogen ring and the substituted.hydrogen, it must get you super loaded” surely that’s gibberish, but you get my point. I don’t have a chemistry degree, can someone explain how to appraise a molecule? Also when people make analogues of a drug, how do they know what to change? And what do the numbers mean fbefore a chemical name, like 3, 4 poly methyl anhydrous methylated cathinone? I made that up btw lol, oris that a real drug?...

 

[negrogesic]

They are able to do this by recognizing the structure activity relationship, since the structure of a compound determines its function. You can think of these chemical compounds as keys that unlock (or lock) certain "doors". By recognizing the shape of the key you can guess which door they unlock. I have ok grasp of it for a lay person, for instance I can look at a chemical compound I've never seen before and make an educated guess about its potential psychoactive properties (for instance, "oh that looks likes it might be a sigma agonist" or "that looks like a synthetic cannabinoid" etc). If you look at enough compounds and read about the activity you behind to recognize which shapes serve as a ligand for which target.

Of course this extends to more than just psychoactive drugs, it extends to anything that interacts biologically. So i know much less for instance about the structural activity relationship of antibiotics, although even then our biological systems aren't independent, for example some quinolone antibiotics exert an antigabaergic effect and its no coincidence that they bear a passing resemblance to drugs like methaqualone (which have a progabaergic effect). Essentially methaqualone is a key with a good fit so it fully activates, where as the quinolone antibiotic has a bad but partial fit so it inhibits.

Structure–activity relationship - Wikipedia

en.wikipedia.org

 

[thegreenhand]

I see some posts where a guys like “well it has the nitrogen ring and the substituted.hydrogen, it must get you super loaded” surely that’s gibberish, but you get my point. I don’t have a chemistry degree, can someone explain how to appraise a molecule? Also when people make analogues of a drug, how do they know what to change? And what do the numbers mean fbefore a chemical name, like 3, 4 poly methyl anhydrous methylated cathinone? I made that up btw lol, oris that a real drug?...

try reading the second half of pihkal (free on erowid)

you can see Shulgin's process for doing this. he was pretty much the pioneer of this stuff

 

[thegreenhand]

I think some of these tools are free to use online : http://www.swissdock.ch/

do note that swiss target prediction is a purely data driven/ML tool. it does not necessarily reflect real world binding data

 

[negrogesic]

I think well before looking at shulgins writings (which are more about psychedelic chemistry than anything else anyhow) or fooling around with the swiss target prediction tool, one simply should become familiar with various targets of interests (opioid receptors, monoamine transporters, NMDA or GABA systems etc etc) and the ligands (drugs) that interact with them.

 

[polarthedog]

Ok, that makes since, I’ll start with pikhal, thanks guys

 

[LucidSDreamr]

It needs to share a very similar core structure to known molecules. Take amphetamine for example. Very small substitutions on the aromatic ring (hexagon) part are known to produce still active drugs.

You have to know that you can't make the amine tertiary and fuck with the alpha methyl substitution if you want a meth like drug.

So basically to make such predictions there already needs to be a well established family of compounds based off of a core structure...where we already know what parts you can modify without destroying activity. That's the only way you can look at it and make decent predictions.

Let's say you lived before fentanyl or methadone were discovered but we had the same level of scientific understanding as right now. It would be impossible to simply look at a brand no class of structure like fent and predict it would do the same thing as morphine based opiates.

Now there are computational chemists that use modeling programs of receptor sites and can fit the chemical structure into the model using computer programs and calculate binding energy and interactions between a totally new molecules and the receptor....but this is not just looking at a structure on paper like you are talking about