Skip to main content

Chemical News

The Smallest Drugs

Here is the updated version of the “smallest drugs” collection that I did the other day. Here are the criteria I used: the molecular weight cutoff was set, arbitrarily, at aspirin’s 180. I excluded the inhaled anaesthetics, only allowing things that are oils or solids in their form of use. As a small-molecule organic chemist, I only allowed organic compounds – lithium and so on are for another category. And the hardest one was “Must be in current use across several countries”. That’s another arbitrary cutoff, but it excludes pemoline (176), for example, which has basically been removed from the market. It also gets rid of a lot of historical things like aminorex. That’s not to say that there aren’t some old drugs on the remaining list, but they’re still in there pitching (even sulfanilamide, interestingly). I’m sure I’ve still missed a few.
What can be learned from this exercise? Well, take a look at those structures. There sure are a lot of carboxylic acids and phenols, and a lot more sulfur than we’re used to seeing. And pretty much everything is polar, very polar, which makes sense: if you’re down in this fragment-sized space, you’ve got to be making some strong interactions with biological targets. These are fragments that are also drugs, so fragment-based drug discovery people may find this interesting as the bedrock layer of the whole field.
Some of these are pretty specialized and obscure – you’re only going to see pralidoxime if you have the misfortune to be exposed to nerve gas, for example. But there are some huge, huge compounds on the list, too, gigantic sellers that have changed their whole therapeutic areas and are still in constant use. Metformin alone is a constant rebuke to a lot of our med-chem prejudices: who among us, had we never heard of it, would not have crossed it off our lists of screening hits? So give these small things a chance, and keep an open mind. They’re real, and they can really be drugs.
Smallest drugs final set2b

30 comments on “The Smallest Drugs”

  1. Anonymous says:

    awesome, let’s look at some structures! Phosphates, Michael acceptors, terminal acetylenes, thiols, these little meanies break all the rules!

  2. jungchemiker says:


  3. Derek Freyberg says:

    What’s fascinating to me about the list is, first, just how many compounds there are on it; second, the diversity of structures (though there are a few similar-looking compounds on the list: phenylephrine, metaraminol, norepinephrine, and the interesting opposite-handed isomer pseudoephedrine); and third, how broad the range of diseases treated is. And, though many are old, they’re still coming – tavaborole was discovered only about 10 years ago (judging by first patent filing).

  4. JBK says:

    Two that you missed:
    1) Chloral hydrate MW 165.4
    2) Gamma Hydroxybutanoic acid (GHB): aka 4-hydroxybutanoic acid MW 104.1

  5. Anonymous BMS Researcher says:

    @jungchemiker: Derek specifically said he was excluding lithium and the like from his list because he is a small-molecule organic chemist.
    I also notice he defined “smallest” as that of aspirin — which is everybody’s favorite example of a very widely used drug that would probably not get very far in anybody’s pipeline today.

  6. Iulian says:

    Methenamine is a great example. In my country they don’t sell it in drug stores anymore because it’s too cheap and nobody makes much profit on it. Plus no sales rep will give gifts/convention invites/ etc to MDs who write perscriptions for $0.50 / box drugs.

  7. a. nonymaus says:

    Plenty of other little amines to go on the list like dopamine, propylhexedrine, and phenylephrine.

  8. Watson says:

    Great, now I’m going to have to create a “DL55” filter…

  9. Watson says:

    You may want to double-check the structures for foscarnet and theophylline.

  10. A Nonny Mouse says:

    …… and pralidoxime.
    Also, theobromine is being used for treating persistent cough in Korea and is under development by 2 UK companies for this indication.

  11. Lord Kelvin says:

    I don’t think these compounds have much to teach our world of modern, mechanistic, target-based drug discovery. Other than that maybe the modern approach isn’t necessarily the best one.
    Most of these compounds are either a) of unknown mechanism, b) analogs of low MW receptor or ion-channel ligands like dopamine, histamine, or glutamine c) not protein binders at all (i.e. chelators, radical traps, prodrugs), or d) covalent or suicide inhibitors. Many could only be found by the the target-agnostic phenotypic approaches of the past.
    I haven’t looked at every single Wikipedia entry, but I would guess that only a third of these interact with a specific protein target in a reversible manner, and most of those are either antimetabolites or receptor ligand analogs. By which I mean to say they are the products of analoging around known structures, rather than the results of screening.
    The compounds are interesting as a reminder of the diversity of chemistry that can be safely and effectively administered to the human body. But as a scientist in the lead discovery area, it saddens me a bit to know that compounds like this cannot be discovered in the paradigm that many if not most of us work under.

  12. An Old Chemist says:

    @#1, Anonymous: Terminal acetylenes are common in drugs and drug candidates, Tarceva is just one example. About a thiol group, it exists in Captopril.

  13. darwin says:


  14. darwin says:

    Whoops-just looked it up. Au mw-196. No dice

  15. D. C. Sessions says:

    Nitrous oxide: MW = 54

  16. ROGI says:

    #11. Lord Kelvin
    They work.

  17. cynical1 says:

    Here’s a couple more I thought of that I believe fall under your criteria.
    Pyridoxine is used to treat people who take isoniazid therapy to prevent the associated neuropathy. (And there’s another B vitamin on the list.) MW 169. Granted you can get it over the counter but same with aspirin.
    Urea is available as an injectable diuretic.

  18. Design Monkey says:

    Leafed through old paper handbook Negwer, Organic chemical drugs and their synonyms. Stuff there is organized by chemical formula, not by MW, but anyway lighter ones tend to be in beginning.
    So, additions to list
    pyrazinamide – tuberculostatic, most likely still in use in poorer countries
    omega aminocaproic acid – bleeding control agent
    bemegride – CNS stimulant, barbiturate antidote
    Mecamylamine, pempidine and the likes
    Now adamantanes. If you list memantine, then amantadine and rimantadine both also are quite in use.
    Urotopine too.
    Coramine – maybe not in USA, but still in use in other nooks of world.
    Sodium phenylbutyrate – if you cheat and don’t count sodium in mass.
    Phenibut – soviet answer to baclophen (who needs that chlorine anyway). In use in Russia and probably some former soviet territories.
    Ornithine – urea cycle booster, ammonia scavenger in case of hepatic damage. Putting in list depends on that, if you want count in, that it is usually used as aspartate salt.

  19. KYosce says:

    Diethylcarbamazine. Urea analog that just squeaks in at MW 199.3
    Anthelmintic on the WHO list of essential medicines.

  20. KYosce says:

    Oops. My bad. Misread the cutoff MW.

  21. gippgig says:

    Why are mercaptopurine (152) and thioguanine (167) drawn as a ridiculous tautomer? Also there is a typo in pseudOephadrine (165).

  22. gippgig says:

    Wouldn’t pralidoxime also be used for organophosphate pesticide poisoning?

  23. gippgig says:

    I believe ethanol is used as an antidote for methanol poisoning.
    Do you want to include… dihydrogen monoxide?

  24. Hap says:

    Do counterions and salts count in MW? Mechlorethamine is probably the hydrochloride salt and not the free base (similar problem to the 3-chloropropylamine claimed in a Dr. Reddy’s patent here), and pralidoxime has to have a chloride counterion.

  25. Watson says:

    @Hap That’s an excellent question. If you read through Lipinski’s Ro5 paper, he suggests that counterions are included in the molecular weight. When it comes to the representative set of compounds in his tabulation, it is apparent that salts have been stripped in the calculations. I would take it as a tacit assumption that we are talking about the active agent, and not the particular formulation. That doesn’t settle the issue of tautomers or predominant ionic forms at physiological pH, though.

  26. Thomas says:

    One more that isn’t a drug, but is a widely marketed low-toxicity selective enzyme inhibitor- it’s just that the enzyme being inhibited isn’t human: glyphosate

  27. Guppy says:

    @6 Iulian: Methenamine is also prescribed less because the mechanism gives doctors the jeebies — decomposition to release formaldehyde in the urine — in regards to potential long-term risks. Nobody has ever run a sufficiently rigorous and long-term trial capable of spotting such problems, and I doubt anyone ever will.
    On the other hand, Geriatricians sometimes use it, probably because they’re usually old-school physicians, and because of persistant abx-resistant UTIs, and because their patients won’t live long enough to make carcinogenicity an issue.

  28. cliffintokyo says:

    Looking at this list of small organic molecules drugs makes me think: 1) that there still must be a wide range of other small molecules that could be explored; and 2) that putting one or more fluorine atoms into a lot of these structures would quite likely generate a whole new set of compounds with interesting and/or useful biological properties.
    OK, its not drug design, but it would keep a large synthesis team busy for ages….
    Before anyone suggests what I ought to do, regrettably I have been a virtual organic chemist (but not honorary, alas!) for (too) many years. Sorry, dudes!

  29. Miro Moman says:

    You should publish a minireview on this topic…

Comments are closed.