Search

Chemical structure optimization is a vital part of early drug discovery projects. Starting with compounds that show activity on the target of interest, the chemical structures are subsequently optimized toward a development candidate (DC) molecule with the best chances of clinical success. However, the DCs in the context of such optimization programs, as well as detailed characterization of major limiting factors, have not been investigated in detail so far. Here, we report an analysis of the historical DC molecules at Novartis since 2005 in the context of their optimization projects. Mapping the DCs into their respective chemical optimization series, we find that these tend to be synthesized rather early in a substantial number of cases. Further analysis of structural properties, ADMET, and potency-related readouts revealed that DC compounds tend to be generally significantly smaller, more permeable, and have higher ligand efficiency than other compounds sent to in vivo PK studies, which we also show for compounds from the same chemical series. Although this might seem obvious to most practitioners in medicinal chemistry, for all of these properties, we could show that they tend to evolve in an undesired direction during structure optimization. This highlights the difficulty of successfully translating our knowledge to medicinal chemistry optimizations.

Early in silico assessment of the potential of a series of compounds to deliver a drug is one of the major challenges in computer-assisted drug design. The goal is to identify the right chemical series of compounds out of a large chemical space to then subsequently prioritize the molecules with the highest potential to become a drug. Although multiple approaches to assess compounds have been developed over decades, the quality of these predictors is often not good enough and compounds that agree with the respective estimates are not necessarily druglike. Here, we report a novel deep learning approach that leverages large-scale predictions of ∼100 ADMET assays to assess the potential of a compound to become a relevant drug candidate. The resulting score, which we termed bPK score, substantially outperforms previous approaches and showed strong discriminative performance on data sets where previous approaches did not.

Machine-learning and deep-learning models have been extensively used in cheminformatics to predict molecular properties, to reduce the need for direct measurements, and to accelerate compound prioritization. However, different setups and frameworks and the large number of molecular representations make it difficult to properly evaluate, reproduce, and compare them. Here we present a new PREdictive modeling FramEwoRk for molecular discovery (PREFER), written in Python (version 3.7.7) and based on AutoSklearn (version 0.14.7), that allows comparison between different molecular representations and common machine-learning models. We provide an overview of the design of our framework and show exemplary use cases and results of several representation-model combinations on diverse data sets, both public and in-house. Finally, we discuss the use of PREFER on small data sets. The code of the framework is freely available on GitHub.

The ability to predict chemical reactivity of a molecule is highly desirable in drug discovery, both ex vivo (synthetic route planning, formulation, stability) and in vivo: metabolic reactions determine pharmacodynamics, pharmacokinetics and potential toxic effects, and early assessment of liabilities is vital to reduce attrition rates in later stages of development. Quantum mechanics offer a precise description of the interactions between electrons and orbitals in the breaking and forming of new bonds. Modern algorithms and faster computers have allowed the study of more complex systems in a punctual and accurate fashion, and answers for chemical questions around stability and reactivity can now be provided. Through machine learning, predictive models can be built out of descriptors derived from quantum mechanics and cheminformatics, even in the absence of experimental data to train on. In this article, current progress on computational reactivity prediction is reviewed: applications to problems in drug design, such as modelling of metabolism and covalent inhibition, are highlighted and unmet challenges are posed., (© 2022 Wiley-VCH GmbH.)

The accurate description of protein binding sites is essential to the determination of similarity and the application of machine learning methods to relate the binding sites to observed functions. This work describes CAVIAR, a new open source tool for generating descriptors for binding sites, using protein structures in PDB and mmCIF format as well as trajectory frames from molecular dynamics simulations as input. The applicability of CAVIAR descriptors is showcased by computing machine learning predictions of binding site ligandability. The method can also automatically assign subcavities, even in the absence of a bound ligand. The defined subpockets mimic the empirical definitions used in medicinal chemistry projects. It is shown that the experimental binding affinity scales relatively well with the number of subcavities filled by the ligand, with compounds binding to more than three subcavities having nanomolar or better affinities to the target. The CAVIAR descriptors and methods can be used in any machine learning-based investigations of problems involving binding sites, from protein engineering to hit identification. The full software code is available on GitHub and a conda package is hosted on Anaconda cloud.

The alternative pathway (AP) of the complement system is a key contributor to the pathogenesis of several human diseases including age-related macular degeneration, paroxysmal nocturnal hemoglobinuria (PNH), atypical hemolytic uremic syndrome (aHUS), and various glomerular diseases. The serine protease factor B (FB) is a key node in the AP and is integral to the formation of C3 and C5 convertase. Despite the prominent role of FB in the AP, selective orally bioavailable inhibitors, beyond our own efforts, have not been reported previously. Herein we describe in more detail our efforts to identify FB inhibitors by high-throughput screening (HTS) and leveraging insights from several X-ray cocrystal structures during optimization efforts. This work culminated in the discovery of LNP023 ( 41 ), which is currently being evaluated clinically in several diverse AP mediated indications.

The tricarboxylic acid cycle intermediate succinate is involved in metabolic processes and plays a crucial role in the homeostasis of mitochondrial reactive oxygen species 1 . The receptor responsible for succinate signalling, SUCNR1 (also known as GPR91), is a member of the G-protein-coupled-receptor family 2 and links succinate signalling to renin-induced hypertension, retinal angiogenesis and inflammation 3-5 . Because SUCNR1 senses succinate as an immunological danger signal 6 -which has relevance for diseases including ulcerative colitis, liver fibrosis 7 , diabetes and rheumatoid arthritis 3,8 -it is of interest as a therapeutic target. Here we report the high-resolution crystal structure of rat SUCNR1 in complex with an intracellular binding nanobody in the inactive conformation. Structure-based mutagenesis and radioligand-binding studies, in conjunction with molecular modelling, identified key residues for species-selective antagonist binding and enabled the determination of the high-resolution crystal structure of a humanized rat SUCNR1 in complex with a high-affinity, human-selective antagonist denoted NF-56-EJ40. We anticipate that these structural insights into the architecture of the succinate receptor and its antagonist selectivity will enable structure-based drug discovery and will further help to elucidate the function of SUCNR1 in vitro and in vivo.

The generated database GDB17 enumerates 166.4 billion possible molecules up to 17 atoms of C, N, O, S and halogens following simple chemical stability and synthetic feasibility rules, however medicinal chemistry criteria are not taken into account. Here we applied rules inspired by medicinal chemistry to exclude problematic functional groups and complex molecules from GDB17, and sampled the resulting subset uniformly across molecular size, stereochemistry and polarity to form GDBMedChem as a compact collection of 10 million small molecules. This collection has reduced complexity and better synthetic accessibility than the entire GDB17 but retains higher sp 3 -carbon fraction and natural product likeness scores compared to known drugs. GDBMedChem molecules are more diverse and very different from known molecules in terms of substructures and represent an unprecedented source of diversity for drug design. GDBMedChem is available for 3D-visualization, similarity searching and for download at http://gdb.unibe.ch., (© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.)

Several recent reports have shown that long short-term memory generative neural networks (LSTM) of the type used for grammar learning efficiently learn to write Simplified Molecular Input Line Entry System (SMILES) of druglike compounds when trained with SMILES from a database of bioactive compounds such as ChEMBL and can later produce focused sets upon transfer learning with compounds of specific bioactivity profiles. Here we trained an LSTM using molecules taken either from ChEMBL, DrugBank, commercially available fragments, or from FDB-17 (a database of fragments up to 17 atoms) and performed transfer learning to a single known drug to obtain new analogs of this drug. We found that this approach readily generates hundreds of relevant and diverse new drug analogs and works best with training sets of around 40,000 compounds as simple as commercial fragments. These data suggest that fragment-based LSTM offer a promising method for new molecule generation.

Dysregulation of the alternative complement pathway (AP) predisposes individuals to a number of diseases including paroxysmal nocturnal hemoglobinuria, atypical hemolytic uremic syndrome, and C3 glomerulopathy. Moreover, glomerular Ig deposits can lead to complement-driven nephropathies. Here we describe the discovery of a highly potent, reversible, and selective small-molecule inhibitor of factor B, a serine protease that drives the central amplification loop of the AP. Oral administration of the inhibitor prevents KRN-induced arthritis in mice and is effective upon prophylactic and therapeutic dosing in an experimental model of membranous nephropathy in rats. In addition, inhibition of factor B prevents complement activation in sera from C3 glomerulopathy patients and the hemolysis of human PNH erythrocytes. These data demonstrate the potential therapeutic value of using a factor B inhibitor for systemic treatment of complement-mediated diseases and provide a basis for its clinical development., Competing Interests: Conflict of interest statement: Novartis Pharma AG has filed patent applications on the compounds. A.S., K.A., N.M., H.S., T.E., C.M.A., A.M.S., S.-M.L., M.C., A.L.-E., S.S., G.W., L.P., V.D., T.F., R.G.K., C.Z., E.V., F.S., B.G., P.E., N.H., T.M.S., F.C., U.A.A., B.H., M.M., R.S., C.W., B.J., J.M., S.F., R.H., and J.E. are or were employees of Novartis Pharma AG during this work., (Copyright © 2019 the Author(s). Published by PNAS.)

The design and synthesis of macrocyclic inhibitors of human rhinovirus 3C protease is described. A macrocyclic linkage of the P1 and P3 residues, and the subsequent structure-based optimization of the macrocycle conformation and size led to the identification of a potent biochemical inhibitor 10 with sub-micromolar antiviral activity., (Copyright © 2018 Elsevier Ltd. All rights reserved.)

Will the targets of the future be covered by the compound libraries of today? This communication will cover a critical review of past strategies before turning to a new measure of diversity, protein pockets. A fingerprint descriptor for pockets will be described., (© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.)

Adenoviral infections are associated with a wide range of acute diseases, among which ocular viral conjunctivitis (EKC) and disseminated disease in immunocompromised patients. To date, no approved specific anti-adenoviral drug is available, but there is a growing need for an effective treatment of such infections. The adenoviral protease, adenain, plays a crucial role for the viral lifecycle and thus represents an attractive therapeutic target. Structure-guided design with the objective to depeptidize tetrapeptide nitrile 1 led to the novel chemotype 2. Optimization of scaffold 2 resulted in picomolar adenain inhibitors 3a and 3b. In addition, a complementary series of irreversible vinyl sulfone containing inhibitors were rationally designed, prepared and evaluated against adenoviral protease. High resolution X-ray co-crystal structures of representatives of each series proves the successful design of these inhibitors and provides an excellent basis for future medicinal chemistry optimization of these compounds., (Copyright © 2014 Elsevier Ltd. All rights reserved.)

The cysteine protease adenain is the essential protease of adenovirus and, as such, represents a promising target for the treatment of ocular and other adenoviral infections. Through a concise two-pronged hit discovery approach we identified tetrapeptide nitrile 1 and pyrimidine nitrile 2 as complementary starting points for adenain inhibition. These hits enabled the first high-resolution X-ray cocrystal structures of adenain with inhibitors bound and revealed the binding mode of 1 and 2. The screening hits were optimized by a structure-guided medicinal chemistry strategy into low nanomolar drug-like inhibitors of adenain.

The successful launches of dipeptidyl peptidase IV (DPP IV) inhibitors as oral anti-diabetics warrant and spur the further quest for additional chemical entities in this promising class of therapeutics. Numerous pharmaceutical companies have pursued their proprietary candidates towards the clinic, resulting in a large body of published chemical structures associated with DPP IV. Herein, we report the discovery of a novel chemotype for DPP IV inhibition based on the C-(1-aryl-cyclohexyl)-methylamine scaffold and its optimization to compounds which selectively inhibit DPP IV at low-nM potency and exhibit an excellent oral pharmacokinetic profile in the rat., (Copyright © 2014 Elsevier Ltd. All rights reserved.)

The integrin leukocyte function associated antigen 1 (LFA-1) binds the intercellular adhesion molecule 1 (ICAM-1) by its α(L)-chain inserted domain (I-domain). This interaction plays a key role in cancer and other diseases. We report the structure-based design, small-scale synthesis, and biological activity evaluation of a novel family of LFA-1 antagonists. The design led to the synthesis of a family of highly substituted homochiral pyrrolidines with antiproliferative and antimetastatic activity in a murine model of colon carcinoma, as well as potent antiadhesive properties in several cancer cell lines in the low micromolar range. NMR analysis of their binding to the isolated I-domain shows that they bind to the I-domain allosteric site (IDAS), the binding site of other allosteric LFA-1 inhibitors. These results provide evidence of the potential therapeutic value of a new set of LFA-1 inhibitors, whose further development is facilitated by a synthetic strategy that is versatile and fully stereocontrolled.

Novel deazaxanthine-based DPP-4 inhibitors have been identified that are potent (IC(50) <10nM) and highly selective versus other dipeptidyl peptidases. Their synthesis and SAR are reported, along with initial efforts to improve the PK profile through decoration of the deazaxanthine core. Optimisation of compound 3a resulted in the identification of compound (S)-4i, which displayed an improved in vitro and ADME profile. Further enhancements to the PK profile were possible by changing from the deazahypoxanthine to the deazaxanthine template, culminating in compound 12g, which displayed good ex vivo DPP-4 inhibition and a superior PK profile in rat, suggestive of once daily dosing in man., (Crown Copyright © 2011. Published by Elsevier Ltd. All rights reserved.)

Density functional chemical shielding calculations are reported for methane and hydrogen disulfide dimers. The calculations show that the contributions of disulfide bridges to the chemical shielding of neighboring protons is sizable at distances that are frequently sampled in protein structures. A semiempirical model of the quantum chemical data is developed. It is shown that magnetic anisotropy effects of disulfide are poorly described by the McConnell equation, both qualitatively and quantitatively. In particular, the ratio of magnetic anisotropy contributions to shielding along and perpendicular to the magnetic anisotropy principal axis do not conform to the predictions of the McConnell equation, and magnetic anisotropy effects are not null along the magic angle axis. A sulfur-based model of the magnetic anisotropy of the disulfide is developed and shown to give much better agreement with the quantum chemical data.

There has recently been considerable interest in using NMR spectroscopy to identify ligand binding sites of macromolecules. In particular, a modular approach has been put forward by Fesik et al. (Shuker, S. B.; Hajduk, P. J.; Meadows, R. P.; Fesik, S. W. Science 1996, 274, 1531-1534) in which small ligands that bind to a particular target are identified in a first round of screening and subsequently linked together to form ligands of higher affinity. Similar strategies have also been proposed for in silico drug design, where the binding sites of small chemical groups are identified, and complete ligands are subsequently assembled from different groups that have favorable interactions with the macromolecular target. In this paper, we compare experimental and computational results on a selected target (FKBP12). The binding sites of three small ligands ((2S)1-acetylprolinemethylester, 1-formylpiperidine, 1-piperidinecarboxamide) in FKBP12 were identified independently by NMR and by computational methods. The subsequent comparison of the experimental and computational data showed that the computational method identified and ranked favorably ligand positions that satisfy the experimental NOE constraints.

omega-Conotoxin MVIIA is a 25-residue, disulfide-bridged polypeptide from the venom of the sea snail Conus magus that binds to neuronal N-type calcium channels. It forms a compact folded structure, presenting a loop between Cys8 and Cys15 that contains a set of residues critical for its binding. The loop does not have a unique defined structure, nor is it intrinsically flexible. Broadening of a subset of resonances in the NMR spectrum at low temperature, anomalous temperature dependence of the chemical shifts of some resonances, and exchange contributions to J(0) from (13)C relaxation measurements reveal that conformational exchange affects the residues in this loop. The effects of this exchange on the calculated structure of omega-conotoxin MVIIA are discussed. The exchange appears to be associated with a change in the conformation of the disulfide bridge Cys8-Cys20. The implications for the use of the omega-conotoxins as a scaffold for carrying other functions is discussed.