Skip to main content

Chemical News

Are Medicinal Chemists Taking It Too Easy?

I was speaking to a university audience the other day (over Zoom, of course) and as I often do I mentioned the studies that have looked at what kinds of reactions medicinal chemists actually use. The cliché is that we spend most of our time doing things like metal-catalyzed couplings and amide formation, and well, there’s a reason that got to be such a cliché, because there’s a lot of truth in that.

At the same time, there’s some evidence that innovative drug molecules come with innovative structures, more often than you’d expect by chance. It’s for sure that some of the hottest research areas right now (such as bifunctional protein degraders) can produce some rather off-the-beaten-track structures. So how do we reconcile these? Can we be making innovative drugs using a bunch of boring reactions?

This new paper (open access) says that yes, we sure can. The authors (from AstraZeneca) first note that about a third of all the reactions in AZ’s electronic notebooks are amide couplings, which sounds about right. They assembled two random sets of 10,000 compounds that had been made and screened in at least two assays, with one of them featuring amide formation and the other with it specifically excluded. These sets (Amide Formation and Other Reactions) were then evaluated by various techniques to roughly measure structural complexity, diversity, and novelty, and in addition the targets that they had hit in past AZ screens were examined.

And as it happens, the Amide Formation set had similar, but slightly higher complexity than the Other Reactions set. The two sets were virtually identical in lipophilicity and percent of saturated carbon atoms, but the amide group was slightly higher in molecular weight and the the number of chiral centers. As for molecular diversity, two different measurements broadly agreed: the Other Reactions set covered more diversity space, but the two sets also had significant non-overlapping regions. That is, the Amide Formation set was not just contained inside the larger diversity space carved out by the Other Reactions set, but had space all its own as well. And there was no real difference in novelty between the two sets, as measured by the number of structures that already occurred in databases such as ChEMBL. And when historical assay behavior was examined, the Amide Formation set had more active compounds in it, while the Other Reactions set covered a slightly wider range of assays themselves. But the two sets had a large overlap in the actual targets covered, so there was, in the end, not a significant difference between the two in “target space”.

The authors suggest that one reason that so simple a reaction as amide formation can hold its own (versus so many other possibilities) is that there are more and more unique amines available for such reactions. They looked through the ELNs for one-step amide couplings that made compounds for testing and examined the amines involved. On average, 8,000 different amines were used each year for such reactions, and every year about 2,000 of them were new. The authors:

In practice, building-block availability is one of the main determining factors. If the desired building blocks are unavailable, the chemist is faced with the decision whether to invest in new route development, or to make analogs with established routes, or to avoid making the target molecule at all. Given the uncertain nature of drug design, investing more time and resources in making a compound does not guarantee improved molecular quality. . .

. . .In medicinal chemistry, we have now reached a state where millions of building blocks have previously been engineered and can now be used in molecular design and synthesis. In addition to the increase in the number of new amines, boronic acids have been another fast-expanding reagent class since the introduction of the Suzuki coupling method

That really has been a change over my career. There are just so many more neat little functionalized compounds available now; it’s become an entire business of its own. As the paper notes, you even have setups such as Enamine’s REAL compound set, which is a virtual-but-easily-made collection via mixing and matching their available building blocks. That one would come out to well over a billion compounds if someone placed an order for the whole collection.

And if we can get our work done via such easy reactions – plenty of experience in doing the reactions, relatively easy purifications, existing scaleup expertise, and so on – then why shouldn’t we? (I should note that the paper under discussion has a lot of good references to past arguments about this issue). That gets to another point I was emphasizing to my university audience: medicinal chemistry is a means to an end. The end, of course, is the discovery of useful drug molecules, and if the synthetic chemistry can (as much as possible) get out of the way of all the other tricky steps in that process, then so much the better.

That’s not to say that we shouldn’t try new reactions or new technologies. Among other things, these can lead to even more new building blocks that can feed into the easy reactions themselves. And God knows, as you develop the SAR of a compound series you may find yourself unavoidably being pushed into difficult chemistry, where you will need all the help you can get and throwing amide couplings and Suzukis at the problem will avail you not. No, we definitely need our skills and our imaginations – but we need them for the times we need them, and when we don’t need them we should speed drug discovery along with the best tools we have for it. To paraphrase Einstein about physical theories, a synthetic route should be as simple as possible, but not any simpler. Getting as much done as you can with the easy methods leaves you more time to tackle the hard stuff. Get flashy only when you have to.

31 comments on “Are Medicinal Chemists Taking It Too Easy?”

  1. neo says:

    Has anyone tried to describe the diversity and prevalence of bond forming reactions involved in Life? Seems amide bond formation is probably pretty high up that list too, and makes for an interesting case study in what you can accomplish with a relatively small number of chemical transformations.

  2. Adam says:

    It may not just be because chemists tend to like easy reactions like amide couplings- if there’s an amide in your molecule (as is often) – why wouldn’t you set up 1000 reactions at once using automated synthesis? It’s not because you are lazy! If that’s the case, of course it outweighs the more complex reactions you need to carefully set up in a fume hood.

    1. sgcox says:

      Nature has created billions of our amazing bodies by amide couplings.
      There is a good reason we should use it.

  3. Albert says:

    Good process chemists would say the same – use the simplest reactions you can get away with. The target is to make the molecule cheaply and withing given timelines. Novelty of the synthesis is only a bonus.

    1. RueBaby says:

      Untrue!! Analogs made with visible light are more equal. Bonus points if you run in library format. Double bonus if they’re all just assay artefacts

  4. Veritas says:

    The amide coupling reaction has served Med Chemists well throughout its history, and it continues to do so. The development of an immense number of new amines means that as long as there are newer amines available, then the applications for amides will also increase as well.

  5. RB Woodward says:

    Go ahead, prepare hundreds of amides using simple coupling. Now, reduce the amides to basic amines using LAH or BH3. Don’t be afraid…it’s only a reduction step. Then, you’ll have a collection of diverse and druglike amines to complement your collection of amides. Good for you. Good for your project. If you haven’t been preparing basic amines, purifying the freebase using silica gel columns (mobile phase: methylene chloride, methanol/NH3), and routinely preparing the HCl salt forms…then you’re asleep at the wheel.

  6. Some idiot says:

    When I started reading the post, my first thought was “well, yes, but the available diversity in amines, acids, boronics etc is pretty substantial, so it makes sense…”.

    And mirroring this in your last paragraph: when a case arises where good new methodology is developed (necessity being the mother of invention; I would guess that observation and serendipity is probably the father…!), I would eat my beloved hat if a large and diverse range of substrates for these transformations did not become rapidly available…!

  7. I want to make medicine for all of us says:

    Hi Derek (and others),

    I’ve asked you this before, in person, as a young hopeful for a career in medicinal chemistry, but its something that continues to vex me: why exactly does pharma search for these R1 style hardcore synthesis types considering the reality of practical synthetic medicinal chemistry? Especially considering the state of tools like Reaxys, Scifinder, and the growing interest in autonomous retrosynthesis technologies?

    I’ve worked for a few years in pharma/biotech in a variety of synthesis roles before beginning my PhD. I was originally trying to head down the total synthesis pathway, but have now diverted to a more computational or (gasp) medicinal chemistry style PhD. I recognize this will likely make my future employment in industry extremely difficult. But in all honesty, between the three companies I’ve worked at, I had never felt that my lack of synthesis experience was holding me back (especially after the first two years 40 hr+/week work experience).

    It’s not that I think I’m super bright (I’m not), but mostly that I’m not sure how developing a new Ir photocat cross coupling, or Baran style decarboxylative carboxylations and so on was going to prepare me for medicinal chemistry beyond teaching me general intuition, technical lab skills, and perseverance. Those skills have also been partially taught to me through my synthetic roles so far (I was certainly not taught photocat in undergrad and have made extensive use of it since) and I felt as though I could have continued to develop them via industry (every company had some sort of weekly synthesis journal club). However, I had seen truly bright chemists with MSc degrees/10 years+ of experience being held in stagnant roles (which ultimately convinced me to leave & get the PhD).

    Even more strangely to me was the fact that at a certain level of experience, these highly skilled synthesis PhD were moved out of the lab anyway. More and more chemistry seemed to be outsourced to CROs, and so on. Of course, these senior chemists are great at directing chemistry but they were now 10-15 years removed from their PhDs in an esoteric field of catalysis or niche total synthesis. I’d argue they had more experience in industrial chemistry than academic chemistry at that point. But maybe I am missing something?

    1. Denovo says:

      Hi, I’m not Derek and would be interested to hear his answer. I am however a big pharma medicinal chemist so I thought I would offer my view. Total synthesis, methodology and all that good stuff has a lasting effect on how you view molecules. Conformation, pka, synthetic complexity and fluency in looking at structures all flow from those experiences. Perseverance and a fearsome work ethic are also par for the course. That said, I’ve met equally good medicinal chemists with more of a computational or chem bio background. The game is undoubtedly changing, and I feel like the age of synthesis jocks running the show is slowly but surely coming to an end.

    2. Macro says:

      And this is precisely the comment I wanted to make.
      In my opinion the point is that chemistry depts are led by chemistry heads who usually went down the total synthesis/ org synth methodology route before finding a place in pharma and having to learn the bio side. Now they believe everyone should follow the same path. There’s a concept that you can teach an org synt chemist biology but you can’t teach a biochemist proper org synt methodology. Sure – as mentioned above it teaches you a a way to look at chemistry, perseverance and such (though the latter is usually a given after most academic backgrounds).

      But the truth is that more often than not, special total synthesis superpowers are really close to useless. Sure, you can get into a challenging lead-op program, which your company wants to invest in. Otherwise you can use these powers for chemistry challenge/journal club sessions if anyone wants to organize them.

      My experience in one of the biggest pharma giants out there is exactly as yours. Not only bringing challenging structural sketches to life is often strongly discouraged, especially if you fail with the initial route (it’s just not worth the effort). I even witnessed library enrichment survey sessions, where suggestions were made on which scaffold families to buy in order to grow the already huge internal screening libraries. The result was very surprising to me then as the flatland widened with every choice. To my very junior question on discarding yet another sp3 molecule I got a very simple answer – would you like to have a lead op program on something like that? True – it would certainly be a costly challenge.

      And yet – all the medchem postdocs can stay for a one-two year internship only to be replaced by total synthesis PhDs.

      1. Mister B. says:

        Just sayin’, but unpublished / unfinished total synthesis don’t bring the chemists far, especially in Big Pharma.

        That’s OK to glorify this subfield of organic synthesis, but for one hired because of his background, countless are left behind with the same training but no publication in the end.

    3. c says:

      FWIW there is a similar academic/industry dichotomy of thought in software engineering.

      A huge majority of software engineering comes down to complexity management. Meanwhile, computer science degrees focus very often on computational theory and algorithm development. The algorithm experts are needed somewhere in the stack, but most programmers are implementing very “simple” stuff in comparison.

      The benefit of a “proper” CS background is similar to the benefit of a “proper” synthetic background: the domain experts know what’s “hard” and what isn’t.

      Idea illustrated:

      (btw this doesn’t address the idea that maybe we have too many of these high-level thinkers in both fields, though)

    4. Marcus Theory says:

      Having seen a number of candidates comes through for entry level roles at the big pharma company at which I work, I think those who went through a medicinal chemistry PhD program are at a disadvantage for three reasons.

      1) Academic labs have a *lot* fewer resources than pharma companies, so the problems that the academics focus on are necessarily smaller — at best, a small but vital part of a larger puzzle (think Cravatt and RASG12C); at worst, an exercise in futility as the lab chases potency but ignores all the other key parameters of drug discovery. Often this makes the candidate seminar narrow and uninteresting compared to the development of a new methodology or the completion of a total synthesis.

      2) Maybe this is just a corollary to point 1, but I think a lot of med chem shops in academia instill a dogmatic view of medicinal chemistry. e.g. “All you need is my PI’s MPO score and you’re all set.” This rubs people the wrong way and can make such candidates sound out of their depth during interviews with longtime medicinal chemists who have seen many kinds of problems. This isn’t to say that dogma isn’t a problem in big pharma. It is. Everyone has their own little mental list of rules that can’t be broken. But it is dangerous to go into an interview with a dogmatic view of the world, having not really begun one’s career in earnest.

      3) The biggie. Most companies want new PhD hires to hit the ground running making molecules. In a twist of painful irony, when we see a bunch of amide couplings and sp2-sp2 cross-couplings in a candidate slide deck, we question whether the skills are there. Those are “easy” reactions — what will the candidate do to troubleshoot something unexpected? The interviewers typically prefer to see a problem-solving seminar where the problems being solved are synthetic, not biochemical or biological.

      That’s been my experience, anyhow. I want to add that I’m skeptical of the method we use to interview candidates. I think the actual job (and its many different kinds of roles) of medicinal chemists is so different from any academic experience that you really only find successful med chemists on the job, after some on-the-job training and real life experience. But you have to have some criteria for hiring, so many people default to the classic organic chemistry job talk + technical interview. BTW I think a whole different set of criteria apply to people with a computational background, and typically they are applying to a very different job req.

      I’d also like to add real quick that I do not believe that automated retrosynthesis programs, or Reaxys/SciFinder, are in a place where you can take a relatively inexperienced chemist and get a good synthesis out. One day, yes. Not today. There is still a lot of chaff in with the wheat and it’s important to have the experience necessary to eyeball a route and say “there will be a regioselectivity issue here; Reaction X never works well; this route is two steps longer but it is much more likely to work first try.”

      1. Derek Lowe says:

        These are excellent points, and I’ve had some very similar experiences. Good advice!

    5. Sanchise says:

      You ask why pharma prefers people with total synthesis, or to a lesser extent methodology training over medicinal chemistry training?

      As an entry level-hire PhD discovery chemist, your job first and foremost is to make compounds for the testing assays. In addition, you’re going be leading/mentoring either in house BS/MS associates (becoming more and more of a rarity nowadays) or outsourced CRO chemists. Even for the most simple & straightforward series, you need to be able to make compounds quickly and troubleshoot issues quickly. While total synthesis does not rule the roost like it does in the 1990s and early 2000s, it still provides the best training for chemists. One needs breadth and depth in knowing functional group transformations, regio, chemo, stereo & enantioselectivity issues, conformational analysis, etc. We need to see some evidence that when issues arise in a synthesis, you will be able to troubleshoot them and develop alternative chemistry in order to synthesize the desired target. You tried to epimerize an alcohol using a Mitsunobu reaction and that did not work as planned. What now? You also want your guidance of internal/CRO associates/chemists to be more than telling people to look things up in SciFinder or Reaxys.

      Medicinal chemistry PhDs by and large create out chemists with inferior synthetic skills relative to total synthesis PhDs. Most of the chemistry performed in a medicinal chemistry PhD program is simple and straightforward, and focuses upon a particular core. One may claim that they are exploring SAR, but usually they are creating analogs with the chemistry that they can perform upon that particular core – and not truly exploring SAR as one would in industry. The best time to learn to troubleshoot reactions, learn a great deal of functional group transformations and how to overcome synthetic challenges is in graduate school. By the time one gets to a post-doc or is applying to industry jobs – if that strong synthetic foundation is not there, it’s too late. You won’t be able to develop that working in industry and keep your head above water. Is this always the case with medicinal chemistry PhDs? It is not always true, but – 97 times out of 100 the synthetic skills are just not there in a medicinal chemistry PhD.

      The second issue at hand – you probably saw this if you worked big pharma and the larger biotechs – is that the chemist hires come from certain groups. Once these entry level hires become group leader or project leader, they tend to hire from similar groups. For example, there are an awful lot of people at Pfizer who worked for Corey. There are an awful lot of people at Merck who worked for MacMillan or Dave Evans. If you don’t come from this caliber of a group, you are going to have a hard time getting a job as a chemist in pharma or big biotech.

      1. Lambchops says:

        “By the time one gets to a post-doc or is applying to industry jobs – if that strong synthetic foundation is not there, it’s too late.”

        Something about this statement sticks in the craw a bit for me. I think it’s perhaps because it says something more about the attitude of the employers rather than the perceived deficiencies of the prospective employees.

        I think it’s partially because, having moved from chemistry to another field, I had to learn some entirely new disciplines from pretty much a standing start. Am I ever going to be a renowned expert in them? No. Can I do the nitty gritty bits of technical analysis/work that are the nuts and bolts of the areas? No. But, can I find a niche where I add value? Yes. Do people give me the support needed to develop further? Yes.

        So as a synthetic methodology PhD who initially went into medicinally chemistry I just find the idea that it’s too difficult to develop people who aren’t even completely new to the field and have many of the relevant foundations (if somewhat more narrowly focussed than desired) – when other fields offer people the chance to do so with only a very limited array of on the job experience.

        Don’t get me wrong I can understand why a hiring manager would rather go for someone who already has all the skills in place rather than needs development, and there’s certainly no shortage of total synthesis or synthetic methodology PhDs to choose from. Just think it’s sad that there still seems to be a lack of willingness to develop people who don’t have the “right” type of background.

  8. Barry says:

    The vertebrate adaptive immune response is our teacher. When we need to cover diversity space w/ antibodies or TCRs, we do it with proteins–i.e. by making amide bonds. Where that lesson is weaker is when we impose restrictions on molecular weight and membrane permeability. Sure, we make lots of amides. And we also spend man-years making bio-isosteres of amides trying to get from lead to drug.

  9. cb says:

    Organic synthesis is not the most difficult part if medicinal chemists want to create excellent NCEs for drug discovery. It is most difficult to create a new molecule which has the following properties: good affinity for the target, good selectivity, good solubility, good chemical stability, good transport across membranes, good stability against many metabolic enzymes, moderate plasma protein binding, poor affinity for efflux transporters, sufficient exposure in target tissue, no accumulation in tissues or lysosomes, no toxic or immunogenic side effects and “safe” metabolites and……. The medicinal chemists should have a good feeling to find the best compromise in property space, because optimizing one property most of the time is on cost of another one. If one would know which molecule is required in this game and there is a patent and a billion dollar market I am pretty sure it can be synthesized and produced at large scale. So please do not deselect compounds in hit- or lead optimization because they would be more difficult to synthesize, whereas its properties as mentioned above are more promising. Perhaps we will see less amides 😉

  10. chemist with no hands says:

    It is quite evident that often you don’t need to make structurally or synthetically complicated compounds to achieve reasonable good control of your target’s functional response. The rest of drug-like attributes named above are also reasonably well understood, and in no way a major challenge. I guess that’s one of the advantages of having a huge drug-like chemical space to work with, the order of 10exp(60) molecules, as I recall.

    The classic definition of the medicinal chemistry role in the pharma industry (of any size) used to be “design and synthesis of drug candidates”.

    The synthesis part has aggressively been moving towards CRO. Why? Because to make amides (or Suzuki couplings, etc) they are more cost-efficient. You can keep a few chemists in-house for the more challenging syntheses. Many don’t, and opt for the 100% virtual organization.

    The design part is how some fortunate chemists have managed to keep their jobs and helped discover good drugs. You can call them medicinal chemists, chemical biologists, designers or whatever. They don’t make the compounds – just cleverly conceive the structure of the right molecules.

    This makes sense: the bottle neck in drug discovery is the accurate prediction of clinical efficacy. Chemists can play a key role there, but it is just not even close to showing up with the “powder in a bottle”.

    The problem is that most organizations (not all) do not understand this complementary (to biologists and other non-chemists team members) and important nature of the chemists’ input. Many organizations have a blind spot, some by design, and consider chemists as unable to provide valuable input other than the solids for testing. Such a shame…

    But some don’t, and they appreciate the value that a solid medicinal chemist can bring to their teams. Good for them!

  11. Peter S. Shenkin says:

    When I worked for Schrodinger, I once visited a small biotech that was pursuing a single target that the founder was an expert on. I presented a core-hopping technique that I thought might lead them to explore more chemical space. They asked whether it included assessment of synthesizability. At that time it did not. They said, “That’s good, because we wouldn’t believe it anyway!”

    That view might no longer be valid, but then they said, “If computational chemistry gets good enough to actually predict binding free energies, we would spend a lot of time trying to synthesize what even looks like a hopeless structure, because we are tied to this one target and we don’t have any other option.”

    That was a useful perspective, although it fails to account for follow-on uncertainties, such as toxicity and promiscuous binding.

  12. Kaleberg says:

    There was a time when a good chunk of chemistry was about searching for new elements. That era is long gone. Chemists have a good, useful list of elements. It seems like drug discovery is approaching this stage in the sense of having a large library for banging together small molecule drug candidates. There will be novel structures, but their usual path will involve discovery of a molecule and mechanism in the wild, so there will still be novel syntheses in our future.

    A lot of fields go through transitions like this. It has been a long while since mechanical engineers have discovered a new gear linkage save by parametric variation of existing structures. Older computer programmers sometimes lament that their field has lost something in that programming now is all about library calls, and there is so much less opportunity to build something from raw seething bits. It can feel strange, but it is a sign of how far we have come, not that we have lost our way.

  13. Ben F scrav3 says:

    I need to fire a a postdoc asap-give me. Give me a micah.

  14. Sam Weller says:

    Unless one shows that chemical suppliers also manage to create complex building blocks with simple reactions, all that the paper has shown is that pharma companies rely on others to do the complex reactions for them.

    “Vether it’s worth goin’ through so much, to learn so little, as the charity-boy said ven he got to the end of the alphabet, is a matter o’ taste.”

  15. Matthew Todd says:

    I remember the point being nicely made right at the start of Sharpless’ click paper that we should stop trying to be clever in making C-C bonds and stick instead to C-heteroatom bonds. “..we address here the development of a set of powerful, highly reliable, and selective reactions for the rapid synthesis of useful new compounds and combinatorial libraries through heteroatom links (C-X-C), an approach we call “click chemistry”” (PMID: 11433435). Nice to see the value in being able to search through a lot of electronic lab notebooks, too, for all the paper notebooks enthusiasts out there. Lovely paper Jonas!

  16. If you search for chemical space exploration, number such as 10^67 chemical entities are described (i.e. more than atoms in the universe?). Interestingly, the existing reactions (leading to the billion synthesizable compounds) does not cross match with existing chemical entities found in nature (about 4000 with med chem interest). There is therefore a huge gap to fill before getting very innovative new chemical entities 🙂

    1. Marko says:

      “…does not cross match with existing chemical entities found in nature (about 4000 with med chem interest) ”

      To my untrained ears ( eyes? ) this seems like an important point , with some ramifications on how efforts should be directed to better ensure success in discovering bioactive compounds.

      For example , do we really think only 4000 chemical entities in nature exist ? What if it’s 10,000 or 40,000 , and we just haven’t discovered them all ? Do we know , for instance , every small molecule , peptide , etc. that plays some role in the life cycle of deep-sea animals and microbes ? Or in the incredible diversity of life in places like the Amazon rainforest ? I suspect not.

      Whatever the actual number , crossing these entities against the immeasurable possibilities that might exist by chemical synthesis seems like an obvious first step to avoid wasting a lot of time chasing ghosts.

      1. Barry says:

        since the advent of mAbs, that 4000 number is nonsense (if it ever meant anything). The entire vertebrate Ab repertoire is now of med chem interest. That’s between 10e15-10e18

        1. Marko says:

          ” The entire vertebrate Ab repertoire is now of med chem interest. That’s between 10e15-10e18 ”

          What a ridiculous , straw-man argument. Nobody here is talking about synthesizing , from scratch , 10e15-10e18 antibody molecules.

          The 4,000 figure may be wrong when talking about the universe of biologically-active compounds that synthetic chemists should have an interest in making or modifying , but not for the reason you cite.

  17. Barry says:

    “since the advent of mAbs as therapeutics”

  18. Mol biologist says:

    There is nothing wrong with the work of a chemist, synthesis with a high yield of the final product and the creation of new active compounds is almost an art. In my opinion, the problems are different, biological science has a historical tradition. Pavlov, Mechnikov, and Pirogov laid the cornerstones and understood the fundamental principles of life processes for a decade or even a century ahead. Molecular biology has made its monumental contribution in terms of molecules and chemical reactions. But no one denies that up to the present time there is a gap between how the molecular biologist understands nature in comparison with the classical biologist. Recent examples with PSK9 or PHD hydroxylase inhibitors and many others have disproved expectations of what should be achieved with real results. In my opinion, success will come only if a bridge is created between the old school and new knowledge when biologists answer the questions to which chemists have no answers.

Leave a Reply to chemist with no hands Cancel reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.

This site uses Akismet to reduce spam. Learn how your comment data is processed.