Thanks for the many answers and bringing up more very valuable points for discussion.
I also like the idea of making the selection a peer review issue for the community. And in favor of moving more towards diverse leads (that contain the fragments but have substantial additions), but these are also already being worked on as @mc-robinson wrote.
Few more suggestions for this or maybe follow up rounds (don’t get me wrong, I understand that this whole project is very pressing and it needs engaged people like you to go forward and make quick decisions!)
maybe share your thoughts about how to select respective designs and include the feedback from the community for the decision.
and maybe put out a list of concrete tasks that will help in the selection (with fixed deadlines) that people can self assign them to and work on.
(some of this might already exist and I simply missed it)
Many people are happy to help and it might be even more rewarding if one sees that the work and time put into it, is well spend because it’s taken into account for the next steps.
Thank you for your message and efforts you are putting into this project.
It would be great to see the hypothesis “live tracking” tool. This, along with an open peer review process of the molecular designs to a broader (expert) scientific community will add a lot of value.
We are not far from a great collaborative scientific discovery.
selection_joost4.csv (1.8 KB)
Hi all,
Attached my contribution. A bit of support for earlier selections and things I think are great. In general I think we should go for a covalent binder, and within that strategy I think there are three tactics:
Keep the fragments small and try to find the best warhead for covalent binding. Then, if a drug-like warhead is found to fit, expand using structure-based design.
Keep the chloroacetamide warhead for now and expand using linking fragments, with the aid of rounds of crystallography. Worry about the exact warhead later
Keep the inhibitor relatively large to guarantee at least some affinity. On that basis, try playing with warheads, then expand (actually a mix of 1 +2, but this was the tactic used to develop acalabrutinib)
I’ve included a comments column to indicate in which tactic each molecule fits
Hi, I agree with a lot of the commenters here. I believe it’s better to have 5 compounds that take you five steps to synthesize rather than 25 compounds that take you 2 steps.
I agree that we should be prioritising compounds with a realistic chance of getting what we want rather than just making a lot of compounds because we can. It looks a reasonably large pocket so we probably need to be aiming to fill all parts of it if we can?
Also, it would be great to have some guidance about structural features that would usually be considered undesirable but might be acceptable in this context. Is it okay to design a compound with one or more unresolved chiral centres? With a Michael acceptor? With an exposed OH that might get glucaronidated? Obviously it may be that nobody knows the answers to these questions at the moment, but if we do know them then we can focus efforts accordingly.
Hi everyone, here my top ten…
All of them docked in autodock with good scores, ADME properties in range for Lipinski´s rule of five. If needed i could share this data too…
I tried to select most diverse submissions from each other, according to replies I´ve seen in the post.
Best wishes, Christian.
To my mind compounds containing covalent fragments are the most promising. Also, the molecules must be synthetically accessible and must not contain toxic fragments.
This is a very cool project in today’s realities, but when you look at many structures, you involuntarily recall the “Mixer Eyes” meme. Unfortunately, one gets the impression, that many authors have’t seen the active site at all, or have no idea what poisons they publish.
Thank you everyone for the submissions so far, this is very helpful!
Sorry I was offline for most of yesterday, but @andrea, I think your post is very important. We definitely want everyone to feel that the work they put in is being valued – and the last thing we would want is for people to feel they are speaking into the void.
Having said that, still trying to figure out this whole collaborative drug discovery paradigm – while trying to move very quickly. Coming from open source software, it’s a difficult balance between:
(1) Try to respond to every feature request and have no coherent product
(2) Have a single person saying exactly what the product will be, and people can implement it if they want.
Trying to work towards the happy medium for drug discovery, wherever that is…
As a side note, we now have a few people in the background making sure we have a more well-defined process for the community, will share when finished
And thanks agian to everyone who has submitted suggestions so far!
Hi @JHullaert, thanks for the input. I’m sympathetic to your view, though will perhaps push back slightly – while open to input/pushback in return from you and others.
Since we are dealing with a fragment screen – many of the binders are quite weak. Since the binding data really isn’t there yet, conventional wisdom may be to start with small changes and build in potency from the beginning as you try to get binding data on fragment like structures.
We are obviously a bit more bullish since the active site is so well populated with structures and are encouraging larger designs. However, as designs get more complex, time+cost go up obviously, but the ability to predict outcomes also becomes seemingly more dubious as we venture further from the known data.
So I guess the quasi-answer is to currently try both in parallel, i.e. lots of small changes and fewer, more complex molecules. Then hopefully a bigger molecule binds, and you can take the data from smaller fragments to adjust potency.
I have to say I’m a bit surprised at the pushback on ordering lots of molecules already in Enamine REAL / 2-3 step simple syntheses, since ordering a screen from a REAL-like database seems to be a common first step in many projects. Here we have just crowdsourced the picking of compounds, and gone ahead if the molecule was already in REAL
Top ten submission.csv (560 Bytes)
I chose these two different series. I guess that 5ec serie could be a covalent inhibitor due to the potential formation of disulfide bond, which could be an alternative to the chloroacetamide serie. I think that it could be useful to explore this molecular space also because the synthesis could be quite easy.
I chose 995 serie because is quite related to the previous one and could be obtained with only one synthesis step more.
Moreover these two series have been also well ranked in the pharmacophore preliminary check. I guess that in this phase, wide molecular diversity on warheads should be explored within relatively simple molecules.
I’d humbly suggest, the first hurdle is to find activity; if this is succeeds then the whole project will get an even bigger boost, if indeed this is possible! . Some/a lot of these things can be fixed later (prodrugs, warhead optimisation, toxicophore substitution, bioisosteres etc… but there’s no harm in considering problematic Herg pharmacophores, nitro groups, mutagenic anilines, etc at the earliest possible stage!
there are even ways of favorably using metabolism as in e.g. the unmasking of a SH in thiophenes as in plavix etc… but these wouldn’t show up in the current biochemical assays.
With two crystal structures of Mpro covalently bound to ketoamide warheads, and both compounds have submicromolar IC50 and micromolar EC50 against this protease, I do not understand why no one is paying attention to the submissions that containing this promising warhead. Yes, those are large compounds, but are already backed up by available structure and assay data.
@JSPEN I agree, some of approved drugs as clonazepam, tolcapone, nitrofurantoin, etc… have nitro group in their structure. Maybe should be good to consider the submissions with good affinity with the main protease and then evaluate toxicity risks.
A lot of the ordered compounds are very interesting and diverse. My comment was fueled by the surprise to see a very strong focus on close analogues of the simple urea fragments in the list of ordered compounds.
My point of view is probably too simplistic for this large project and as non-expert in fragment based discovery I should be more carefull giving my opinion on things.
@JHullaert , please keep giving feedback – the point of this type of open collaboration is not to be shy about giving advice/ asking questions. We are all learning and adjusting as we go. I for example, was all ready to triage about 50-100 submitted designs, but then rapidly had to adjust cloud resources and change strategy to deal with the 3500 submissions.
Dear organizer,
Here come my top10 submission.
From all the submitted compound, I performed :
Molecular descriptors are calculated in order to visualize the chemical space based on druggable phys-chem properties (Hbond-don, Acc, NRotBond…).
Molecular fingerprints are calculated with obabel and compared to the other in order to build a NxN Tanimoto matrix used to visualize the chemical diversity of the submitted molecules.
Visual selection of several compounds pre-selected with the help of Phys-Chem and Chemistry spaces. (Using my personal submission as reference, and also trying to select a set of molecules that explore different part of the chemical space).
At the end I selected 3 covalent and 7 non covalent ligand.
Hi Lyna. I did some work on the PDBs with the ketoamide inhibitors. They nicely overlay with some fragments. See JOO-IND-926. And the screenshots I posted on the entry at the forum - best Joost
Top10.csv (620 Bytes)
Attached are my top 10.
I have excluded compounds with unprotected amides as although they are key for recognition by the protease they suffer from significant metabolic instability especially when dosed orally. This could be fixed in lead optimization but could also lead to significant loss of activity. Use of bioisosteres or steric shielding would likely provide metabolic stability.
I also excluded compounds with likely high lipophilicity/lipophobicity.
Some of the compounds possess significant pi-electron systems that would likely interfere with fluorometric assay, applying a PAINS screen would be highly recommended.
Thanks JoostU for your response. For keta-amide warhead, based on Hilgenfeld group’s extensive SAR, it is best to keep lactam at P1 and a small leu or cyclopropyl group for P2. So a fast and promising route to improve IC50 from micromolar to nM is to optimize P1’ and P3 sites. This design idea can take full advantage of crystalized covalent fragments as many of them overlay with the keto-amide warhead and P1’ benzyl in the x-ray complex. In addition, we also screened 470 drug fragments at P1’ and P3 sites using SILCS. Thus the designed molecules are quite big and complex because peptidomimetics do not necessarily follow the rule of 5 YUN-WES-64c. Appreciate any feedback. Thanks!