Skip to main content

In Silico

Virtual Covalent Screening

Covalent drugs have been a big item in R&D over the last few years, and I wrote here about covalent fragments. The whole topic of reactive groups in small molecules and their interaction with living systems and biomolecules is a complicated one, with many interesting twists.
Now the Shoichet group at UCSF has what could be a useful computational approach to the field. They’re reporting “DOCKovalent”, a virtual screening platform for covalent inhibitors, and illustrate it with examples of cyanoacrylamides and boronic acids across several different enzymes. The calculations are based on the angles needed for an electrophile to react with a residue like cysteine – these reactions, as organic chemists know, can be rather constrained in what approaches the reacting partners have to take. In solution you can see stereoelectronic effects that arise from the structures of the small molecules, but most reactions can find a way. When this process is taking place in the clefts of a protein, though, the number of feasible approaches can get cut down considerably. I’m generally pretty hard to convince when it comes to virtual screening, but the number of constraints needed here gives me more hope than usual for meaningful results.
Running their program retrospectively, with known covalent inhibitors of various enzymes versus decoy molecules, showed that the virtual screening did (in most cases) give hit sets that were well enriched in the real binders. Like any such technique, it’s going to fail sometimes, both on individual compounds and on whole runs, but the Shoichet group has made the tool available on the web for anyone to try for free. Here’s the link, and I applaud them for putting it out there. The appeal of computational approaches has always been the low barrier to entry, and this lowers it even more. That’s the hazard of computational approaches, too, of course, but the results from a virtual screen like this should at least be easy to subject to a real-world test.

15 comments on “Virtual Covalent Screening”

  1. crock says:

    Like most computational techniques, and as you properly mention, they only used this technique after the fact. No real predictions made. What a crock! Wait for a real-world test before spreading it.

  2. anonao says:

    I don’t have to all the paper but the abstract says:
    We apply this method prospectively to discover reversible covalent fragments that target distinct protein nucleophiles, including the catalytic ​serine of ​AmpC β-lactamase and noncatalytic ​cysteines in ​RSK2, ​MSK1 and ​JAK3 kinases. We identify submicromolar to low-nanomolar hits with high ligand efficiency, cellular activity and selectivity, including what are to our knowledge the first reported reversible covalent inhibitors of ​JAK3.
    It has the word prospectively not retrospectively, plus data to back it up.
    So crock you may want to read a bit more before commenting.

  3. Wavefunction says:

    It’s a nice paper. I think there’s a lot of promise in developing both computational and experimental approaches for studying *reversible* covalent inhibitors and I can only see the field growing in the future. Schrodinger has also come up with a useful approach that’s worth exploring (linked in handle)

  4. Anonymous says:

    Boronic acids bind to glycans on the surfaces of cells. How would those molecules even make it inside of a cell? The efficiency may be awful after most of those molecules are sequestered by the cellular surface interacting with god knows what.
    BAD-Lectins: Boronic Acid-Decorated Lectins with Enhanced Binding Affinity for the Selective Enrichment of Glycoproteins
    Anal Chem. 2013 Sep 3;85(17):8268-76

  5. Pete says:

    We published an article (linked as URL for this commnent) on automated molecule editing last year. We showed how our molecule editor MUDO (a SMIRKS-based re-write of Leatherface that some of your readers may be familiar with) could be used to build models of covalently bound complexes. MUDO was used to add hydrogen sulfide across the olefinic bond of a vinyl sulfone (K777) which generated two steroisomers which were docked into a cruzain structure in which the catalytic cysteine had been mutated to glycine. We then used MUDO to delete the ligand thiol and form a covalent bond between the catalytic cysteine sulfur atom and the appropriate carbon atom of the ligand. MUDO was built with the OEChem toolkit (OpenEye) and source and some input/output examples are provided in the supplemental material. I should point out that our principal reason for writing the article was to illustrate applications (e.g. tautomers, ionisation, states, matched molecular pair analysis) for automated molecule editing in pharmaceutical design and so the covalent docking is just a short section in the paper.

  6. anon says:

    Sorry – off topic, but is anyone having trouble getting to from IE 11.0.14. We recently upgraded at work and now when I visit I get:
    “Fatal error: Call to undefined function: str_split() in /home/corante/public_html/pipeline/connect.php(1) : regexp code(1) : eval()’d code(1) : regexp code on line 1”
    I thought pipeline was down for a week!

  7. Hap says:

    I had the same problem – not home uses a much older IE version and it fails in exactly the same way. I asked Chemjobber, and once he said it worked for him, I tried Firefox and it worked OK. For a while, I thought Corante had finally failed completely.

  8. Dr Octopus says:

    Looks like an update to Internet Explorer caused the problem. We just swapped to a different browser for looking at ITP.

  9. Nir London says:

    Thanks for the write-up Derek. I should disclaim that I’m a lead author on this paper.
    As noted, we did undertake prospective application of this approach and discovered new ligands against three different protein targets. In full dose-response, the best of these new inhibitors ranged from 10 to 40 nM IC50. The structures of several were determined by x-ray crystallography. These inhibitors were dissimilar to previously known inhibitors of the targeted AmpC, JAK3 and RSK2 kinases, and were shown to be active in cellular assays.
    With regards to the cellular efficiency of boronic acids, while some boronic acids can indeed interact with glycans, this doesn’t preclude them as useful cellular agents. We showed in this study that the new AmpC inhibitors were able to potentiate cefotaxime in clinically isolated resistant bacteria (with no effect in the absence of cefotaxime). Velcade – an approved drug for multiple myeloma is another example of a useful boronic acid based inhibitor.
    I should say that this project emerged from a close collaboration between the Shoichet & Taunton labs, who did all the new synthesis and the kinase biology.

  10. Nick K says:

    #6: Same problem here. The Pipeline loads perfectly with Firefox, but not at all from IE, instead giving me the same error message that you see.

  11. Crock says:

    everyone knows that no one watches the computationalists. Were the prospective predictions deposited anywhere or are we suppose to trust another “too good be true” result? Was the code even archived and available before the experiments were undertaken? Of course not.

  12. Anonymous says:

    @11 Crock, are you alleging straight-up fraud? Usually, around here, when people accuse wet lab papers of fraudulence, they point to some evidence such as photoshopped figures. Are you saying we should assume good faith from biologists or chemists and assume malfeasance from computational work?

  13. crocked Crock says:

    Crock might just be bitter at losing his/her job to a computational biologist or chemist. Clearly he/she did not bother to read the paper, only Derek’s (incomplete) summary. The authors first employed the technique retrospectively to answer the question if it is good enough to capture known binders. Then the authors used the technique prospectively to find new binders, which were then further validated. True, the labs studied classes of proteins that they are experts in, and in remains to be seen if it will be truly effective across other classes of proteins in the proteome. That does not mean it should be dismissed. We need new ways to discover small molecules apart from just trying to make a bunch of compounds and screening them.

  14. Non-crocked says:

    It’s ok crock, you’ll find a new job soon. Although it might take longer if you keep on jumping to wrong conclusions without actually reading a paper.

  15. According to me Biosimilars and high-throughput screening are the fast and reliable ways for Drug Discovery and even Dendrimers technique as well

Comments are closed.