SCUBIDOO: A Large yet Screenable and Easily Searchable Database of Computationally Created Chemical Compounds Optimized toward High Likelihood of Synthetic Tractability

CHIPMUNK: A Virtual Synthesizable Small-Molecule Library for Medicinal Chemistry, Exploitable for Protein-Protein Interaction Modulators

A common issue during drug design and development is the discovery of novel scaffolds for protein targets. On the one hand the chemical space of purchasable compounds is rather limited; on the other hand artificially generated molecules suffer from a grave lack of accessibility in practice. Therefore, we generated a novel virtual library of small molecules which are synthesizable from purchasable educts, called CHIPMUNK (CHemically feasible In silico Public Molecular UNiverse Knowledge base). Altogether, CHIPMUNK covers over 95 million compounds and encompasses regions of the chemical space that are not covered by existing databases. The coverage of CHIPMUNK exceeds the chemical space spanned by the Lipinski rule of five to foster the exploration of novel and difficult target classes. The analysis of the generated property space reveals that CHIPMUNK is well suited for the design of protein-protein interaction inhibitors (PPIIs). Furthermore, a recently developed structural clustering algorithm (StruClus) for big data was used to partition the sub-libraries into meaningful subsets and assist scientists to process the large amount of data. These clustered subsets also contain the target space based on ChEMBL data which was included during clustering.

Generating Focused Molecule Libraries for Drug Discovery with Recurrent Neural Networks

In de novo drug design, computational strategies are used to generate novel molecules with good affinity to the desired biological target. In this work, we show that recurrent neural networks can be trained as generative models for molecular structures, similar to statistical language models in natural language processing. We demonstrate that the properties of the generated molecules correlate very well with the properties of the molecules used to train the model. In order to enrich libraries with molecules active toward a given biological target, we propose to fine-tune the model with small sets of molecules, which are known to be active against that target. Against Staphylococcus aureus, the model reproduced 14% of 6051 hold-out test molecules that medicinal chemists designed, whereas against Plasmodium falciparum (Malaria), it reproduced 28% of 1240 test molecules. When coupled with a scoring function, our model can perform the complete de novo drug design cycle to generate large sets of novel molecules for drug discovery.

Binding-Site Compatible Fragment Growing Applied to the Design of β2-Adrenergic Receptor Ligands

Fragment-based drug discovery is intimately linked to fragment extension approaches that can be accelerated using software for de novo design. Although computers allow for the facile generation of millions of suggestions, synthetic feasibility is however often neglected. In this study we computationally extended, chemically synthesized, and experimentally assayed new ligands for the β₂-adrenergic receptor (β₂AR) by growing fragment-sized ligands. In order to address the synthetic tractability issue, our in silico workflow aims at derivatized products based on robust organic reactions. The study started from the predicted binding modes of five fragments. We suggested a total of eight diverse extensions that were easily synthesized, and further assays showed that four products had an improved affinity (up to 40-fold) compared to their respective initial fragment. The described workflow, which we call "growing via merging" and for which the key tools are available online, can improve early fragment-based drug discovery projects, making it a useful creative tool for medicinal chemists during structure-activity relationship (SAR) studies.

Virtual Exploration of the Ring Systems Chemical Universe

Here, we explore the chemical space of all virtually possible organic molecules focusing on ring systems, which represent the cyclic cores of organic molecules obtained by removing all acyclic bonds and converting all remaining atoms to carbon. This approach circumvents the combinatorial explosion encountered when enumerating the molecules themselves. We report the chemical universe database GDB4c containing 916 130 ring systems up to four saturated or aromatic rings and maximum ring size of 14 atoms and GDB4c3D containing the corresponding 6 555 929 stereoisomers. Almost all (98.6%) of these ring systems are unknown and represent chiral 3D-shaped macrocycles containing small rings and quaternary centers reminiscent of polycyclic natural products. We envision that GDB4c can serve to select new ring systems from which to design analogs of such natural products. The database is available for download at www.gdb.unibe.ch together with interactive visualization and search tools as a resource for molecular design.

Molecular Connectivity Predefines Polypharmacology: Aliphatic Rings, Chirality, and sp3 Centers Enhance Target Selectivity

Dark chemical matter compounds are small molecules that have been recently identified as highly potent and selective hits. For this reason, they constitute a promising class of possible candidates in the process of drug discovery and raise the interest of the scientific community. To this purpose, Wassermann et al. (2015) have described the application of 2D descriptors to characterize dark chemical matter. However, their definition was based on the number of reported positive assays rather than the number of known targets. As there might be multiple assays for one single target, the number of assays does not fully describe target selectivity. Here, we propose an alternative classification of active molecules that is based on the number of known targets. We cluster molecules in four classes: black, gray, and white compounds are active on one, two to four, and more than four targets respectively, whilst inactive compounds are found to be inactive in the considered assays. In this study, black and inactive compounds are found to have not only higher solubility, but also a higher number of chiral centers, sp3 carbon atoms and aliphatic rings. On the contrary, white compounds contain a higher number of double bonds and fused aromatic rings. Therefore, the design of a screening compound library should consider these molecular properties in order to achieve target selectivity or polypharmacology. Furthermore, analysis of four main target classes (GPCRs, kinases, proteases, and ion channels) shows that GPCR ligands are more selective than the other classes, as the number of black compounds is higher in this target superfamily. On the other side, ligands that hit kinases, proteases, and ion channels bind to GPCRs more likely than to other target classes. Consequently, depending on the target protein family, appropriate screening libraries can be designed in order to minimize the likelihood of unwanted side effects early in the drug discovery process. Additionally, synergistic effects may be obtained by library design toward polypharmacology.

The octet rule in chemical space: generating virtual molecules

We present a generator of virtual molecules that selects valid chemistry on the basis of the octet rule. Also, we introduce a mesomer group key that allows a fast detection of duplicates in the generated structures. Compared to existing approaches, our model is simpler and faster, generates new chemistry and avoids invalid chemistry. Its versatility is illustrated by the correct generation of molecules containing third-row elements and a surprisingly adept handling of complex boron chemistry. Without any empirical parameters, our model is designed to be valid also in unexplored regions of chemical space. One first unexpected finding is the high prevalence of dipolar structures among generated molecules.

Neural-Symbolic Machine Learning for Retrosynthesis and Reaction Prediction

Reaction prediction and retrosynthesis are the cornerstones of organic chemistry. Rule-based expert systems have been the most widespread approach to computationally solve these two related challenges to date. However, reaction rules often fail because they ignore the molecular context, which leads to reactivity conflicts. Herein, we report that deep neural networks can learn to resolve reactivity conflicts and to prioritize the most suitable transformation rules. We show that by training our model on 3.5 million reactions taken from the collective published knowledge of the entire discipline of chemistry, our model exhibits a top10-accuracy of 95 % in retrosynthesis and 97 % for reaction prediction on a validation set of almost 1 million reactions.

Expanding Synthesizable Space of Disubstituted 1,2,4-Oxadiazoles

One-pot synthesis of 3,5-disubstituted 1,2,4-oxadiazoles from carboxylic acids and nitriles was optimized to parallel chemistry. The method was validated on a 141 member library; the desired products were recovered with a high success rate and in moderate yields. Practical application of the approach was demonstrated in the synthesis of bioactive compound pifexole and agonists of free fatty acid receptor 1. A library of 4 948 100 synthesizable drug-like 3,5-disubstituted 1,2,4-oxadiazoles was enumerated based on the method and available validated reagents.

SwissSimilarity: A Web Tool for Low to Ultra High Throughput Ligand-Based Virtual Screening

SwissSimilarity is a new web tool for rapid ligand-based virtual screening of small to unprecedented ultralarge libraries of small molecules. Screenable compounds include drugs, bioactive and commercial molecules, as well as 205 million of virtual compounds readily synthesizable from commercially available synthetic reagents. Predictions can be carried out on-the-fly using six different screening approaches, including 2D molecular fingerprints as well as superpositional and fast nonsuperpositional 3D similarity methodologies. SwissSimilarity is part of a large initiative of the SIB Swiss Institute of Bioinformatics to provide online tools for computer-aided drug design, such as SwissDock, SwissBioisostere or SwissTargetPrediction with which it can interoperate, and is linked to other well-established online tools and databases. User interface and backend have been designed for simplicity and ease of use, to provide proficient virtual screening capabilities to specialists and nonexperts in the field. SwissSimilarity is accessible free of charge or login at http://www.swisssimilarity.ch .