10 February 2025

Flatland: still a nice place to be

In 2009 we highlighted a paper reporting that approved drugs have a higher fraction of sp3-hybridized carbon atoms than discovery-phase compounds. Perhaps focusing on molecules with a high Fsp3, the ratio of sp3-hybridzed carbons to total carbons, would lead to greater success. Or not: a new analysis in Nat. Rev. Chem. by Ian Churcher (Janus Drug Discovery), Stuart Newbold, and Christopher Murray (both at Astex) finds that the relationship has not held up.
 
The 2009 “Escape from Flatland” paper was widely discussed at conferences and has been cited more than 3000 times. But according to the authors of the new study, most of these citations are from papers describing new synthetic methodologies rather than from papers discussing medicinal chemistry.
 
And not all the attention has been positive. As we noted in 2013, Pete Kenny and Carlos Montanari reanalyzed the data and found that an apparent correlation between Fsp3 and solubility disappeared when plotting all discrete data points instead of binned data.
 
More recently, we highlighted a paper that found no significant difference between the shapeliness of drugs, as assessed by their principal moment of inertia (PMI), and the shapeliness of small molecules in the ZINC database.
 
The new paper looks at Fsp3 values for drugs approved during various time periods. Among 980 drugs approved up to 2009, the average Fsp3 was 0.458. However, of the 431 drugs approved after 2009, the average Fsp3 has dropped to 0.392. The researchers speculate that this (statistically significant) decrease may be due to an increase in the number of kinase inhibitors, which are usually highly aromatic, as well as an increase in the use of metal-catalyzed cross coupling reactions.
 
In my analysis of the 2009 paper, I asked whether higher Fsp3 ratios would lead to lower hit rates, and indeed this seems to be the case, as shown in a paper we discussed in 2020. Thus, if you pursue difficult targets, you may increase your chances of finding hits by screening molecules with lower Fsp3 ratios. Also, multiple studies, including one published just last month, have found no correlation between the shapeliness of a fragment (as defined by deviation from planarity, or DFP) and the shapeliness of the resulting lead, so there appears to be no penalty to starting with a flattish fragment. 
 
The researchers conclude that their “analysis of drug development trends over the last 15 years suggests that Fsp3 may not have been a useful metric to optimize.” Importantly, the supplementary information includes a list of >1400 approved drugs and >1500 investigational drugs along with associated properties, so you can do your own analyses.
 
In the end, generalizations will only get you so far, and may even lead you astray. At least for now, there are few shortcuts in the long slog of experimental studies necessary to discover a drug.

03 February 2025

Stitching together fragments with Fragmenstein

As we noted just last week, crystallography has unleashed a torrent of protein-ligand complexes, especially fragments. Historically a single structure might be used for fragment growing, but so many structures present an embarrassment of riches, with sometimes dozens of fragments that bind in the same region. Merging or linking these fragments can be done manually, as seen here and here, but how to do so when the binding modes are partially overlapping is not always intuitive. In a new open-access J. Cheminform. paper, Matteo Ferla and colleagues at University of Oxford and elsewhere describe an open-source solution called Fragmenstein.
 
We briefly described Fragmenstein in 2023, where it was used to combine pairs of low-affinity fragments bound to the Nsp3 macrodomain of SARS-CoV-2 to generate sub-micromolar inhibitors. The current paper describes the platform in detail.
 
Fragmenstein starts by taking two (or more) structures of fragments bound to a protein and virtually combining them. This is done by collapsing rings to their individual centroids, stitching these together along with their substituents, and then re-expanding the ring(s) so the substituents will be close to where they were in the initial fragments. This process produces a surprising array of molecules beyond the obvious. For example, if one fragment contains a phenyl ring and the other fragment contains a furan ring, the stitched molecule might just contain the phenyl (if the two rings bind in nearly the same position), a benzofuran (merging the rings), a phenyl ring linked to a furan by one or more atoms, or even a spiro compound if the rings are perpendicular to one another.
 
In silico approaches sometimes suggest molecules that are synthetically challenging to make, but Fragmenstein can also be used to find purchasable analogs.
 
Next, the new molecules are energetically minimized, first by themselves and then while docked into the protein. In contrast to other docking programs, which might allow molecules to sample thousands of different conformations and sites in a protein, Fragmenstein maintains the new molecule in a similar position and orientation to the initial fragments, with the assumption that these have already identified energetically favorable interactions.
 
The researchers successfully applied Fragmenstein retrospectively to several targets. The COVID Moonshot (which we discussed here) crowd-sourced molecule ideas for the SARS-CoV-2 main protease based on structures of bound fragments. Of 87 ligands that had been crystallographically characterized and were designed based on two fragments, Fragmenstein successfully (RMSD < 2 Å) predicted the binding mode for 69%.
 
Fragmenstein can even be used for covalent ligands, as shown for the target NUDT7, which we wrote about here. Merging two fragments led to compound NUDT7-COV-1, and the RMSD between the Fragmenstein model and the crystal structure was an impressive 0.28 Å.
 
Of course, as the researchers acknowledge, the number of possible analogs might be daunting, and deciding which to make or buy is not necessarily straightforward. Also, Fragmenstein assumes that the fragments themselves are making productive interactions with the protein, which may not be the case, as we suggested here. Still, the tool is open-source and worth trying, especially if you are swimming in crystal structures.