Docking studies on MurC ligase

[jhjensen2]

In preparation for our study on using a genetic algorithm to find good binders for MurC (more on that later), Casper Steinmann (@cstein) has docked 250K molecules from the ZINC database and 10K molecules from Enamines Diversity Discovery Set.

The docking is performed with Glide using the SP scoring methodology using the 6X9F crystal structure. The 1000 best binders from each set are then redocked using the more accurate XP scoring methodology. Below we show the 100 molecules with the best (lowest) docking scores.

For comparison AZ8074 has a docking score of -7.2 using the same methodology.

Here are screenshots of the top 20 molecules for Enamine and ZINC, but all the data can be accessed in this notebook.

Enamine

ZINC

[dkoes]

If you are testing compounds, would it be possible to test some of the compounds I proposed (see Jul 6 message)? I'd be happy to order the MCULE compounds (send me shipping info via email). The ChemSpace compound is available from ENAMINE (Z1980956983) so it would be more cost effective to add it to an existing order.

[jhjensen2]

@KatoLeonard But those are from the screen of the smaller DDS. We have subsequently screened the larger HLL set and found molecules with better docking scores (see this post from July 15). @cstein has rescored them using LigPrep and we'll share the results (including docking poses) here within the next few days. Stay tuned

[jhjensen2]

Turns out that we get worse docking scores with LigPrep, so I suggest selecting molecules from this batch. @cstein is working on getting the docking poses. Time permitting, he'll also dock them with DOCKER.

[cstein]

I have (finally) extracted the 20 best poses of our docking studies on the HLL database. They are available for download as a .zip file with both SP and XP results. You can match the docking scores for the SP results with our (@jhjensen2) July 15th post (https://github.com/opensourceantibiotics/murligase/issues/46#issuecomment-880537392).

Any feedback is welcome of course!

[jhjensen2]

Here are our results for the Enamine Hit Locator library of 234K molecules. The docking is performed with Glide using the SP scoring methodology using the 6X9F crystal structure. The 1000 best binders from each set are then redocked using the more accurate XP scoring methodology. The scoring is done with LigPrep.

Compounds with more than 5 rotateable bonds (bad for accumulation) and docking scores > -7.0 (AZ8074 has a docking score of -7.2) are removed as are molecules with DOCK scores > -50. All compounds have a globularity < 0.25, which is good for accumulation. But some of them have computed logP values >3.5 (Z57909504, Z2038227779, and Z57907808).

If our results for the D4 dopamine receptor are transferable, then there's a 60% chance these molecules are "good" binders.

@cstein can you upload the docking poses for these molecules?

The csv file can be found here HLL_top_ligprep.csv

[cstein]

The poseviews for the above ligands are extracted and available here (compare the titles in Maestro with the title id in the .csv file)

Download link: https://www.dropbox.com/s/llc4ofxqahgr8ei/dock_sorted_poses.zip?dl=0

[jhjensen2]

Here are the results from a genetic algorithm search for primary amines (reminder: primary amines are less likely to be pumped out of the cell).

You can see the first 10 molecules above (the rest are in this csv file: amines_20211126.csv), with the name, docking score, and number of synthetic steps predicted by Postera’s Manifold retrosynthesis program. The molecules are sorted in order of number of synthesis steps and then by docking scores.

Things to help guide molecule selection:

• Look at the docking poses (@cstein will post these soon).  Does the ligand make the same kinds of interactions (e.g. H-bonds) with the protein as AZ5595 and AZ8074 do? Is the protonation state realistic?
• Consider the ease of synthesis. To see Manifold’s retrosynthesis predictions you can paste the smiles strings from the csv file in their website https://postera.ai/manifold/. (You have to make an account but it’s free). Do the suggestions make sense? Are the building blocks available at a reasonable price? How easy would it be to make derivatives? You can see an example Manifold output below
• Consider chemical diversity. For example several molecules have a quinolinone group. Maybe just try a 1-2 of those?

[cstein]

Here are the link for a .zip-file with two pose-viewer (rigid and confgen) files for the results that @jhjensen2 posted a few days ago. When you load them up, they are in the same order as in the .csv files that Jan posted.

Link for download: https://www.dropbox.com/s/d7leyvs7yul3opc/amines_GB-GA_poses_20211129.zip?dl=0

[jhjensen2]

I have made a few slides for tomorrow's meeting. I've gone through and picked seven molecules for possible synthesis, based on their poses, ease of synthesis, etc. Amine_poses.pdf