GDB-17: 166 billion fragments and counting

How many possible fragments are there? Jean-Louis Reymond and colleagues at the University of Berne have been trying to answer this question computationally by enumerating all stable molecules from first principles. In their previous effort they found nearly a billion molecules with up to 13 atoms. In a new paper published in J. Chem. Inf. Model. they have now extended this analysis to molecules containing up to 17 carbon, oxygen, nitrogen, sulfur, and halogen atoms. There are 166,443,860,262 of them.

What do they look like? Before addressing that question, it is worth noting that this set of molecules—dubbed the GDB-17—is not exhaustive. The researchers intentionally excluded many potentially unstable moieties. Most of these are probably best ignored, though doing so does leave out functionalities found in some drugs, such as hemiaminal ethers (acyclovir), sulfoxides (omeprazole), and some non-aromatic double bonds (cyclosporine). In fact, more than 40% of similarly-sized molecules in PubChem (ie, they’ve actually been synthesized) are not represented in GDB-17.

But even looking at the PubChem molecules that do show up in GDB-17, there are dramatic differences between existing molecules and enumerated possibilities. For example, a huge fraction of the GDB-17 set contains 3- or 4-membered rings. Aromatic rings are surprisingly rare, at only 0.8%, compared with roughly a third of similar-sized molecules in PubChem. On the other hand, 57% of the GDB-17 molecules contain nonaromatic heterocycles, compared with just 12% in PubChem; these may be particularly attractive in terms of drug-like properties.

Consistent with their reduced aromaticity, the GDB-17 molecules also contain many more stereocenters: on average more than 6 per molecule compared with just 2 for PubChem. With all these stereocenters, it’s inevitable that the molecules are more three-dimensional than those in PubChem. By the same token, though, synthetic accessibility is likely to be a challenge.

In general the GDB-17 molecules are also considerably more polar than known molecules of similar size, with more than half of them having ClogP ≤ 0. This could in part be due to the fact that it’s often tough to purify molecules that are too soluble in water.

There’s a lot of other fun stuff here: for example, almost half a billion isomers of procaine! And for the synthetic chemists in the audience, this work illuminates vast fields of uncharted chemical territory, just waiting to be explored.

Fragments of the future - part 3 (977 million and counting)

A few weeks ago we gave a passing nod to work by Jean-Louis Reymond, who with colleagues enumerated all possible compounds with up to 11 C, F, N, and O atoms. In a new JACS Communication with Lorenz Blum, he has now expanded this analysis to molecules containing up to 13 non-hydrogen atoms.

The new dataset, GDB-13, contains 977,468,314 molecules containing carbon, oxygen, and nitrogen atoms (as well as hydrogens, of course). Unlike its predecessor it excludes fluorine, but it happily adds chlorine as an aromatic substituent as well as sulfur in heterocycles or in sulfones, sulfonamides, or thioureas. To speed calculation (which still required the equivalent of 4.5 years of CPU time), a few other simplifications were made to limit the number of heteroatoms in a given structure.

The resulting collection, while huge, is thus obviously incomplete: about two-thirds of 619,675 molecules that contain up to 13 atoms and are reported in a variety of databases do not appear in GDB-13. And GDB-13 has many unconventional structures – over half of the molecules contain one or more three- or four-membered rings.

Still, there is lots of neat stuff here: for example, 804,153 structural isomers of aspirin, and 18,371,393 structural isomers of mexiletine! And since 45.1% of the new molecules are rule-of-three compliant, there are hundreds of millions of virgin fragments just waiting to be made – and tested.

Fragments of the future - part 2 (The intersection of chemical and biological space)

In a previous post about heterocycles that appear chemically feasible but have not been reported, we wondered whether these molecules would show biological activity. The structure of biologically relevant chemical space – that fraction of possible molecules that will exhibit some biological effect – is of great interest, but as yet unknown. Brian Shoichet and coworkers at UCSF have just published a thought-provoking analysis in Nature Chemical Biology that is also relevant to developing new fragment libraries. 

The researchers ask why it is that HTS collections of a million or so compounds, vanishingly small in comparison to the roughly 1,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000 possible small drug-sized molecules, nonetheless so often succeed in identifying hits. A lovely paper by Tobias Fink and Jean-Louis Reymond had previously computationally enumerated all possible compounds with up to 11 C, N, O, and F atoms. Of these 26,429,328 molecules, 25,810 are commercially available. 

Shoichet and colleagues compared the structures of these compounds with the structures of metabolites and natural products (all of which have by definition been processed by at least one protein) and found that the commercially available compounds were much more similar to natural products and metabolites than were non-commercially available compounds. Indeed, the more similar a molecule is to a known natural product or metabolite, the more likely that it is available for purchase; 2918 of the commercially available compounds are in fact natural products or metabolites. 

The bias also increases exponentially with molecular size: a random 11-atom commercial compound is almost 1000-times more likely to resemble a natural product or metabolite than is a non-commercially available molecule, whereas the bias is only about 2-fold for 6-atom molecules. Similar results were observed with other libraries. 

The authors conclude that: A major reason why the screening of synthetic compounds ever finds notable hits is that our libraries are biased toward the sort of molecules that proteins have evolved to recognize.  

This resemblance is reasonable. After all, most commercially available compounds are ultimately derived from naturally occurring starting materials, so their similarity to natural products isn’t surprising. Moreover, historically much of chemistry was devoted to natural product synthesis, so many of the intermediates built up over the years resemble natural products. And of course, once you learn how to do chemistry on one moiety, you will tend to stick with it unless you have a good reason to do otherwise; each heterocycle behaves (often frustratingly) differently, so if a natural-product-like molecule does the job, why look for trouble? 

But does this “biogenic bias” mean that the rest of chemical space is a biological desert? Not necessarily. I can imagine at least two alternative models of chemical-biological diversity space. 

Let’s call one model “lamp posts in dark fields.” Consider a vast field of some crop that can only be harvested by night. There are lights scattered haphazardly throughout the field. One might expect that the crops immediately under the lamp posts would be harvested more intensively than crops in darker parts of the field, even if other areas are equally productive. In this scenario, the lamp posts reveal natural products and similar molecules, but much – or even most – of (unlit) chemical space may also be biologically active, it just hasn’t been sampled yet. 

Another possibility is the “oil-field model.” As we are all too aware, petroleum is distributed very unevenly across the globe. In some areas, such as Texas, oil was easy to find and easy to extract. In others, such as the deep ocean or the high arctic, oil is harder to find and more technically demanding to access. In this scenario, there are vast pockets of chemical space that are relevant to biology, they just haven’t been identified (let alone accessed) yet. 

These are fun, speculative questions, but the paper provides some practical data. Specifically, 83% of core ring scaffolds found in natural products are absent from commercial libraries. In fragment or lead-sized molecules of MW < 350 with less than three stereocenters, 1891 rings scaffolds found in natural products are not commercially available. These could be useful additions to fragment libraries, and the paper lists 18 examples. 

In fact, at least one company, deCODE, is explicitly enriching its fragment collection with molecules based on natural products and metabolites. This might be a good strategy. After all, even if the “lamp posts in dark fields” model is correct, there are plenty of brightly illuminated, unharvested chemotypes. At least for now, picking these may be more productive than venturing into the twilight-zone of uncharted chemical space.