Skip to Content

Drug Development

How to Be a Good Medicinal Chemist

Longtime medicinal chemist Mark Murcko has a Perspective article out in J. Med. Chem. on “What Makes a Great Medicinal Chemist“. As he makes clear from the beginning, if you’ve been doing this stuff for a while, you’ve likely heard many of these recommendations before. But it’s useful for people starting out, and it’s also useful for organization to think about what they’re encouraging (and what they’re not!)

There are several characteristics common to good scientists in general – paraphrasing, we have: (1) intellectual curiosity, (2) ability to focus on the problem at hand, (3) pragmatic approach to problem-solving, (4) respect for data, (5) attention to detail, (6) sense of urgency, (7) awareness of other work in the field, (8) openness to new technologies (once they’ve proved their utility!), (9) willingness to challenge assumptions and ditch old ones, (10) enjoyment of their work, (11) awareness of the limits of their knowledge, (12) resiliency, (13) effective at communication, (14) working well with team members (there are exceptions!), (15) not constantly seeking credit, and (16) looking for mentors early on, and serving as one later.

Past that, he has traits that are more specific to medicinal chemists: (17) never losing sight of the final drug profile, (18) the ability to come up with unusual structural ideas when needed, (19) attention to more than just potency, (20) three-dimensional thinking, (21) awareness of the need for different chemical series/scaffolds, (22) awareness of the competition, but not being scared of IP space, (23) never giving on validated targets, since they’re rare, (24) a deep understanding of biology, (25) decisiveness about what the next steps in a project should be, (26) being unafraid of tough chemistry (in a good cause), (27) a willingness to repurpose anything (in a good cause), and (28) awareness of the history of the field and its current state.

That’s quite a list! And it should be clear at the outset that you’re not going to find people – or not too many – who have all those qualities turned up to 11. But they’re all useful and important, and it’s worth looking over them to see what your strengths are (and especially what you might need to shore up a bit).

A few comments on specific items: I think that (3), (9), and (27) are aspects of the same approach to applied research – that is, a willingness to go with what works, even if it isn’t your idea, even if it disproves your favorite assumptions, even if it seems weird or even if it seems obvious or old. Things that work, that really work, are the ace cards in the deck. Solid experimental evidence brushes aside feelings, theories, and assumptions (quality 4 above ), and you have to be ready to come to terms with it and figure out where it’s telling you to go next (quality 25). A corollary is that if you’re running experiments whose results are not going to be solid or important enough for you to base such decisions on, then why? We’ve all sat in meetings where someone outlines an experiment that (if it works) means the project goes on as planned. But if it doesn’t work? Then. . .the project goes on as planned anyway. It’s hard to see how that’s not, at some level, a waste of effort.

This also has a connection with (11), knowing what you don’t know. There are mistakes at both ends of the scale (there always are), and in this case it’s being too tentative or being too confident. Chemists, in my experience, are more prone to the latter, the mistake the Greeks called hubris. Anyone who’s honest with themselves about biomedical research has to admit that most of the time, about most things, we can only aspire to the level of actually knowing jack diddley. What really causes Alzheimer’s? What are the specific neuronal processes involved in consciousness? What’s the specific molecular process you’d need to hit to stop multiple sclerosis or lupus? On a more immediate level, why does Compound Series A get into cells while Compound Series B doesn’t? What do I need to do to the lead compound to make it about two times more potent at the target? Or about ten times less potent at that other target it’s hitting? What other targets is the thing hitting, anyway? And so on. Decision-making under these conditions is. . .nontrivial.

That’s why many of the desirable qualities above can in fact work against each other, if taken to extremes or misapplied. Attention to detail is important – but not to the point where you become unable to make a call because you can think of bad consequences every way you turn. Ability to focus is important, but not if it leads to glossing over key experimental data that are trying to tell you something. Mistakes at both ends of the scale!

(24), understanding of biology, deserves some comment. My feeling is that not many people from other fields tend to know much about organic or medicinal chemistry, nor do they seem particularly eager to learn. So we chemists have a real advantage, knowing this stuff already, but you can’t rest on that advantage, either. The point is to start from there and go learn about the other stuff. You can be useful as a chemist (at an entry-level position) without knowing much about the biology involved – just bang out compounds. But you cannot be really good at the job with that approach, nor (as time goes on) will you continue to be as valuable to your employer – or any other employer. The chemistry is just a starting point.

One quality worth noting is (12), resilience. A constant feature of drug research is that most things, most of the time, don’t work that well. Your ideas about improving the compounds? By golly, here come the assay data, and it’s back to the bench we go. You have to be ready for your best ideas, your biggest bolts of inspiration, to land in a tangled heap over and over. It takes getting used to, and some personality types are just not suited to it. Longterm med-chemists exhibit survivorship bias; they’re the ones who have been able to deal, but not everyone can.

I’ve been focusing on individual behavior, but as mentioned above, there’s something to be said about organizations, too. Managers of all types should think about whether their labs, departments, or companies are providing a good home for these behaviors or not. Be honest. I’m not talking about wall posters, e-mail newsletters, or “State of the Company” speeches. Your mission statement might say that you’re committed to innovation or something, but talk is as cheap as it’s ever been. For employees looking at their own management, my advice is always this: don’t pay so much attention to what they say, but rather watch what they do. Don’t read the slogans on the intranet pages – pay attention and watch what behaviors your company actually rewards.

If they’re valuing the sorts of things on Murcko’s list, that’s good news. But if they’re not, they may be rewarding some less worthwhile behaviors, like delivering on deadline even if it means delivering crap, whacking people on the head for coming up with inconvenient data or ideas, advancing projects based more on who proposed them than intrinsic merit, etc. There’s a whole Devil’s Dictionary/Screwtape Letters list of things that good medicinal chemists and scientists don’t do, and you need to be alert to their presence, too. The devil finds work for idle hands, and for empty heads, too.

42 comments on “How to Be a Good Medicinal Chemist”

  1. A Nonny Mouse says:

    Can’t see “luck” anywhere.

    1. anonymous says:

      Good one!

    2. ScientistSailor says:

      If you are all of those things, then you are very lucky…

  2. Marcin says:

    Amen

  3. TMS says:

    Chance does favor the well prepared mind (or something like that)

  4. Chrispy says:

    LPT: Search the DOI: 10.1021/acs.jmedchem.7b01445 in SciHub to avoid the paywall…

    1. David Edwards says:

      I thought Sci-Hub was now dead?

      1. Sbd says:

        About as much as Wikipedia

  5. MoMo says:

    I read it with tears streaming down my eyes-

    1. Isidore says:

      Tears of joy, not doubt.

      1. MoMo says:

        Yes, Tears of Joy! Murcko’s psychoanalysis and list of the 28 Great Traits of a Medicinal Chemist resonates and with many of those I know in this discipline. I would of added openly antagonistic of authoritarian figurines, but still a landmark paper!

  6. Wavefunction says:

    1. Don’t put too much faith in rationality (over-interpreting SAR, extrapolating from small data sets, trusting ‘computational drug design’ and ‘AI’)
    2. Don’t fall for technological solutionism; let your approach be problem-based, not technology-based.

  7. JustAnotherPhD says:

    Perhaps some hiring managers in Big Pharma should take note of this.
    Attend any careers talk in a chemistry department, and you’ll hear a manager from a Pharma company insist that all new hires should be experts in synthesis, ideally from total synthesis backgrounds. Yet those particular skills are essentially absent from Murcko’s criteria. Perhaps this’ll stop recruiters writing off perfectly capable PhDs who happened to do medicinal chemistry/chemical biology instead of total synthesis.

    1. Marcin says:

      It is because the hiring manager wants to do all the thinking and not be threatened in his future career

      1. maybe says:

        or because hiring manager HAS total synthesis background

        1. JustAnotherPhD says:

          Ah yes, hiring someone whose background looks exactly like theirs… even though most of their TS-specific training will be useless in a medicinal chemistry job. That’s definitely not a terrible idea.

    2. cynic says:

      Nope, it won’t. Try to find statistics on how many entry-level med chemists BMS hired not from Baran lab and cry.

      1. East Coast Cynic says:

        Or how many entry level scientists at Merck not from the MacMillan, Buchwald, or Jacobsen groups…

        1. anon says:

          I’d take those over any other average organic chemist.

    3. Anonymoose says:

      (1) intellectual curiosity, (2) ability to focus on the problem at hand, (3) pragmatic approach to problem-solving, (4) respect for data, (5) attention to detail, (6) sense of urgency, (7) awareness of other work in the field, (8) openness to new technologies (once they’ve proved their utility!), (9) willingness to challenge assumptions and ditch old ones, (10) enjoyment of their work, (11) awareness of the limits of their knowledge, (12) resiliency, (13) effective at communication, (14) working well with team members (there are exceptions!), (15) not constantly seeking credit, and (16) looking for mentors early on, and serving as one later.

      Past that, he has traits that are more specific to medicinal chemists: (17) never losing sight of the final drug profile, (18) the ability to come up with unusual structural ideas when needed, (19) attention to more than just potency, (20) three-dimensional thinking, (21) awareness of the need for different chemical series/scaffolds, (22) awareness of the competition, but not being scared of IP space, (23) never giving on validated targets, since they’re rare, (24) a deep understanding of biology, (25) decisiveness about what the next steps in a project should be, (26) being unafraid of tough chemistry (in a good cause), (27) a willingness to repurpose anything (in a good cause), and (28) awareness of the history of the field and its current state

      1. Anonymoose says:

        Of those, at least 20 are developed/required in a synthetic organic background.

  8. Anonymous says:

    I was disappointed to see “data” relegated — or denigrated — to the #4 position. Derek paraphrased Murcko as, “(4) respect for data.” Murcko actually wrote, “Obsessed with data” and expanded that to, “They are obsessed with data. They crave it, agonize over it, and constantly question it.”

    In my experiences, many can be obsessed with data, crave, agonize and sometimes question data but all too frequently have no respect for (good, validated) data. Whether it is chemistry (spectral data inconsistent with the proposed structure; exaggerated yields or purities; etc.) or biology (assays that do not work yet provide lots of “data” for powerpoints) or data from other steps along the path to a drug, I have seen good data ignored or SUPPRESSED in order to feed an ego or the bottom line. That sometimes also means that data known to be bad or unreliable was deliberately elevated in importance and misrepresented. (Started to write some stories … too long, too many, hence deleted.)

    If I had to draw the “interactome” of Murcko’s bullet points, most of them would be strongly influenced by DATA, perhaps as the master controller.

  9. cynical1 says:

    What makes a great medicinal chemist? A job?

    Let’s also admit that being a great medicinal chemist and keeping your job in this industry is basically antithetical at this point. Drug discovery management do not reward the qualities listed above.

  10. Useless Molecule says:

    I am very surprised (and somewhat concerned) that “leadership skills” where left out in Murcko’s analysis. I know several “great” medicinal chemists who possess most, if not all, of the listed criteria but went absolutely nowhere in their career or never produce anything of value because they have zero leadership skills, zero charisma and zero influence. Leadership skills and the power of influencing others is an absolute requirement for success in this business, like it or not.

    1. CB1 says:

      The organization will quickly identify the “leaders” after a couple years, the so called “chosen ones”. I know some at each education levels that possess leadership characteristics, but typically the PhD will grow into a true leadership role. I think we all know people that are just not cut out for leadership. Now if you can’t collaborate these days, you might as well just hang up the lab coat.

    2. lola says:

      @Useless Molecule: I partly agree: leadership is important, but would counter what you say by stating ‘predominantly for leaders’. The world today is obsessed with ‘leadership’, but the idea that everyone has to be a leader is inane – summed up by the old aphorism ‘too many chiefs, and not enough Indians’ as a byword for getting nothing done. A good leader only produces based on the contribution of his/her ‘doers’. Good ‘followship’ is a thing too, and while not everyone should aspire to stay as a good ‘doer’ or ‘follower’, the ones who do the actual work are….the ones who do the actual work. Everyone in an lab knows of an outstanding ‘doer’ who quietly ‘does’, and their praises are not sung enough I believe! Industries like IT consulting firms recognise this, and provide career pathways for outstanding ‘doers’ who don’t aspire to leadership positions, recognising their contribution as critical to achieving any outcomes.

  11. Abbey says:

    I don’t see “PhD from a hardcore synthetic organic chemistry group” on the list; is this an unwritten requirement? Do other types of PhD chemists have no hope of being a medicinal chemist? What about, say, someone in a chemical biology group who just so happens to do organic synthesis almost every day?

    1. CB1 says:

      I had to laugh when I read this post, because it is so obvious how in-bed some organizations are with various research groups. I would think most newly minted PhDs or post-Docs with proven work in synthesis have a fair chance of doing well in Medicinal Chemistry or Process. Chemical Biology is a unique area. You definitely need good chemists to make tool-probes that are usually not trivial but more importantly, the open-mindedness to work through initial murky pharmacological data and constructively be critical of experiments to eventually validate/identify a target.

    2. anonymous says:

      Having worked in chem-bio lab I can assert that those who are trained only in chemical biology don’t have excellent synthetic skills and at the same time their mol-biol/biochemistry techniques are average at best. That is in a broad sense, they do often master their very specific narrow set of experiments to perfection. Their superpower is to come up with crazy multi-step experiments, then to come up with even crazier sets of experiments to validate the first ones, and to connect the dots. No need for complex synthesis, no need to make too big of a library. So assay development or instrument-specific job would probably be a good way to apply those skills, not sure why people are still dreaming of med-chem jobs.

      1. PhDreams says:

        I am on my fourth year into getting my PhD in chem-bio and you are right. I am not well-versed in synthetic chemistry and yes, still working on my biochem assays.
        My goal is to move back to the biopharma after I graduate (I left the industry to pursue my PhD), but now I have to confess feeling extremely insecure about my future back into the job market. The list is beautiful and God, how much I’d love to score high in all of those points, but my biggest score is on the “I work really hard” item. Which now I am not sure how much is valuable!
        What I am learning throughout my PhD? I am a hard working, passionate chemical biologist, and to my disappointment, a very average scientist. Nothing groundbreaking will come out of my thesis and I am not from a well-established and famous lab.
        Damn, this was quite of an unsolicited amount of emotional heart-pour.

  12. Sans sheriff says:

    I thought all u need is helvetica, pastels, and the most nominal photocatalyst

  13. Azetidine says:

    Guess what? There is NO SUCH THING as being a great (or even good) medicinal chemist. The name of the game is getting a drug to market and you will be very lucky if you even contribute to one such drug. Essentially zero chance of getting two or more.

    The rest is just hand-waving. Everything you do short of getting the drug to market is play-acting at *medicinal chemistry* (it’s actually just chemistry).

    How can you say you are any good at the job if you never get to the goal?

  14. Eugene says:

    One through 16 are attributes you want to have no matter what you do and hope for in those around you. The rest are can be re-stated for any endeavour. Most people I have worked with have many of these attributes. What I find is missing is leadership. It has been rare for me to have a leader that just does it as naturally as breathing. Most of the time it is like they are following an instruction manual, but very poorly. Combine the ingredients, 1 unit of Scientist, 2 units of Engineer and 2.5 Software Engineer, stir and out comes product. While they are waiting play with Gant charts and budgets. EOR (end of rant)

    1. anon says:

      To reiterate, it does seem that 1-16 are great qualities to have for problem solving in many highly technical endeavors.

      Isn’t this list a veiled endorsement for scientists and organizations that value problem solving first, rather than value pleasing the boss/superior first?

      1. anon says:

        ..approaches that manifest as two personality types that are frequently at odds

    2. anon says:

      leadership is gained from by studying human nature not bench science. Most of the time, like vigor, it is innate.

  15. Anoin says:

    Isn’t this a bit like how to be a good photographic film developer?

  16. anchor says:

    Too sad that these advises are freely coming when the medicinal chemist jobs went extinct! Wrong advice for wrong time.

  17. anon the II says:

    Maybe “How to have been a Good Medicinal Chemist” would be a better title.

  18. Pfizer Joe says:

    You left out how to manage upwards!

  19. Mine’s a Guinness says:

    Don’t see “likes a beer” in the list Derek? Many of the best medchem ideas, and collaborative networks, cultivated in the bar

    1. Anon says:

      “Many of the best medchem ideas, and collaborative networks, cultivated in the bar”.

      At least they seem like a great idea at the time. 😀

Leave a Reply

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

Time limit is exhausted. Please reload CAPTCHA.