Card Design for chemical mechanism and schemes

Hello everybody, this post may be a bit of help leeching, but to quote bernie “i am once again asking for your […] support”

I am a pharmacy student and for the final exam in medicinal chemistry we have to learn the synthesizing paths of quite a few pharmaceutical ingredients. Following you will find one (larger) example how it is teached to us, and what is expected in the exam. We are asked to know around 60 of such syntheses, some larger some smaller

My idea on how to learn them with Anki would be through Image Occlusion.
The card would start at first position, and everything behind it would be covered. Asked would be the first step, so everything above and below the first arrow and the next structure. This would be repeated until the final structure has been reached. Everytime i would judge myself with the usual grading that Anki provides (easy, again, hard, etc) Below is a graphical representation of my thought. (sorry for my shitty gimp skills…)

Now, i wouldn’t want to learn the whole thing from the begining all the time if i am only struggling with step 5. I also made the experience that the “Hide All, Guess One” Option from Image Occlusion doesn’t really fit this use case, as i would have to guess step 5 without having the structure from the step before, and “Hide One, Guess One” gives too much context. So i kinda imagine a “Guess one after another”

So, has anyone an idea on how to implement this? Or maybe has a totally different idea on how to learn this kind of task? My other thought was maybe to implement every step as a cloze, but that sounds like way more work and possible formatting issues.

Thank you all very much in advance.

It will create just one card though, which you can reveal incrementally.

I’m not sure what the best course of action in this case would be. How fluent are you in organic chemistry and reaction mechanisms, the logic behind synthesis so to be speak? It doesn’t really make sense to really learn this information for real longterm purposes, but Anki can be helpful anyways. I’d probably cram some key information of these steps. Otherwise, depending on how much time you have, I guess I’d try to understand the logic behind each step.

To give an example based on the oseltamivir synthesis you show here. Please note that the scheme you have shown sucks and prevents you from learning these steps via logic. There are multiple errors included, and the stereo-configuration of the final product is NOT shown completely. This wikipedia image is much better:

I also recommend reading the paper the synthesis originates from:

  • Which cheap starting material is close to oseltamivir in structure and could be used?
    shikimic acid

  • Which manipulations do you need to get shikimic acid into oseltamivir?
    ethyl ester on the carboxylic acid, pent-3-yl ether moiety, conversion of the two remaining alcohols into amino functionality, and one of them needs to turn into an amide (chemoselectively).

So the next question would be, in which order could one achieve that? Esterification of the Shikimic acid is an obvious one, you don’t have to worry about the carboxylic acid anymore, and conditions for esterification won’t interfer with the alcohol groups etc.

After esterification, they want to chemoselectively attach the pent-3-yl moiety to that one alcohol. Chemoselectively is reached by cleverly using the stereoconfiguration of shikimic acid, because these two alcohols are adjacent and can form a ketal with cheap pentan-3-one, catalysed by cheap p-TSA. This method of ketal formation and later reduction also retains the stereochemistry.

Now we need to worry about beginning to install the amino groups, while inverting the stereoconfiguration of the two alcohols. The free alcohol is mesylated with MsCl (error in your scheme, there is an O missing…). This is clever because it dictates the stereochemistry of the following intermediates, by making the alcohol with that configuration the leaving group for the later epoxide formation.

But before we do this epoxide stuff, we need to free the alcohol group. This is done by silylation (TMSOTf) and reduction of the ketal to an ether (borane reduction with dimethylsulfide-borane complex). The silylation is done for affecting the chemoselectivity (please don’t ask me why), thus affording the product with the pent-3-yl group majorly at the right place (10:1:1 right product, dihydroxylated, pent-3-yl group at wrong position). We are at 4 now. The following base-catalysed (cheap KHCO3) epoxidation will kick out the mesylated group (easy leaving group), yielding an epoxide with both bonds facing away from you (your scheme has it wrong again!). That epoxide is product 5.

  • Now why do we need an epoxide which is facing in that direction?
    because in the conversion from the later aziridine 7 to 8, the resulting aziridine can only be attacked from below, but the carbon in the middle is protected from the bulky pent-3-yl on the adjacent ether, which results in the right beta-azido amine for chemoselective acetylating that middle amine to an acetyl amide!

All that is basically left is learning the conditions for the steps we left out, that would be epoxide opening by NaN3 (5 → 6a/6b), reduction of the azide (Ph3P and aziridine formation) to 7, and that reduction of the azido-group to amine from 9 - 10 again by a phosphine. The last salt formation from 10 to 11 is trivial lol.

At this point, you may rightfully ask: How am I supposed to learn 60x THIS WALL OF TEXT and instantly go into panic mode. Well, your fault you chose studying pharmacy. Just kidding. But I don’t see any other chance than supplementing rote-memorisation by logic (when rote-memorisation fails), and logic by rote-memorisation (when logic fails). Depending on the time available, damage control considerations, etc., the way I’d (personally!!!) go about this exam if I wanted to ace it with the minimal available time:

  • Do the above analysis for each reaction, to the best of my abilities (without wasting time by looking too deep in to the reasonings if I can’t understand them in a reasonable amount of time) → form Anki cards which ask about the key genius points of the synthesis (aka the ketal formation, using the pent-3-yl group as a protecting group for later chemoselective acylation of that one specific amine to an amide, stereo inversion via epoxide/azide/aziridine SN2 magic). Now how is that better than rote memorisation? If you’re doing it right, understanding those keysteps will result in DA activation and usually very low forgetting rates as opposed to rote memorisation of material you have little understanding of. If you’re emotionally dead and can’t feel dopamine spikes anymore, ask your local doctor for a supplementation with crack cocaine.
  • Learn product - (conditions) - > educt for each reaction, yielding 3 cards each. You can do that with the standard cloze deletion. I also have a card template which also allows you to enter a named reaction and mechanism, but that is probably overkill. If you ask kindly, I could attach it, but honestly, cloze should be enough for this use case.
  • Still do the Image Occlusion like you do them to get an overview of the synthesis
  • Right before the exam, do entire memory dumps. Write down the ENTIRE synthesis on prompt from scratch, without any cues, until you make zero mistakes (you can use Anki for that, but keep in mind that this is a cramming step. Definitely toss those cards after you pass the exam, and don’t sweat about learning this longterm. This should be done in the days right before the exam IMO).
  • Choose your learning material wisely. If the above scheme is how you are supposed to learn it, obviously keep it in mind, but the example you gave is wrong at some points and SHIT for understanding at all the other points. IIRC there are a couple books which explain total synthesis of a lot of medications, but I never really read into those and while they might include the same target molecules, maybe the process is different bla bla. But worth looking into, because learning the logic from there might be easier than digging for the original references and possibly falling into a rabbit hole.

Please note that this is my approach as a chemistry student who never had to learn 60+ total synthesis approaches, just the usual biochem pathways. I am very sure it would work for me (as I have a relatively intuitive grasp of organic chemistry), but your mileage may vary. Good luck!