Enumeration of 166 Billion Organic Small Molecules in the Chemical Universe Database GDB-17

Abstract

Drug molecules consist of a few tens of atoms connected by covalent bonds. How many such molecules are possible in total and what is their structure? This question is of pressing interest in medicinal chemistry to help solve the problems of drug potency, selectivity, and toxicity and reduce attrition rates by pointing to new molecular series. To better define the unknown chemical space, we have enumerated 166.4 billion molecules of up to 17 atoms of C, N, O, S, and halogens forming the chemical universe database GDB-17, covering a size range containing many drugs and typical for lead compounds. GDB-17 contains millions of isomers of known drugs, including analogs with high shape similarity to the parent drug. Compared to known molecules in PubChem, GDB-17 molecules are much richer in nonaromatic heterocycles, quaternary centers, and stereoisomers, densely populate the third dimension in shape space, and represent many more scaffold types.

Communication: Understanding molecular representations in machine learning: The role of uniqueness and target similarity

“…Initial structures for all datasets were drawn from the GDB universe. 32,33 Links to all data sets are available at http://quantum-machine.org.…”

Finding Chemical Structures Corresponding to a Set of Coordinates in Chemical Descriptor Space

“…There were 26.4 million chemical structures in the database (26,434,571 structures), and all the compounds were chemically relevant because they passed filters, which eliminated unstable and hard to synthesize structures. [36] Although Reymond's group also compiled other chemical universe databases (GDB13 [37] and GDB17 [38] ), we chose GDB11 in this analysis because the aforementioned databases were compiled by applying additional filters, which make the contained chemical structures plausible for drug design and lead discovery, such as the element-ratio filters for GDB13 [37] and the no-small-ring filter for graphs having 17 heavy atoms in GDB17 [38] as well as the computational easiness of treating relatively smaller number of chemical structures.…”

Maximizing computational tools for successful drug discovery

“…Despite the seemingly large chemical library produced by these efforts, we are still far from reaching the estimated chemical space of > 10 60 molecules. The first step to scratching the surface for expanding our coverage of the chemical space was carried out by the research group of Reymond in which 166 billion molecules containing up to 17 atoms (i.e., C, N, O, S and halogens) were computationally enumerated [7]. Another approach for increasing the sampled chemical space is to employ molecular fragments where some of which are privileged structures that can bind a wide range of receptor sub-pockets and act as good chemical starting points [8].…”