Biophysics-Guided Lead Discovery of HBV Capsid Assembly Modifiers.

Hepatitis B virus (HBV) is the leading cause of chronic liver pathologies worldwide. HBV nucleocapsid, a key structural component, is formed through the self-assembly of the capsid protein units. Therefore, interfering with the self-assembly process is a promising approach for the development of novel antiviral agents. Applied to HBV, this approach has led to several classes of capsid assembly modulators (CAMs). Here, we report structurally novel CAMs with moderate activity and low toxicity, discovered through a biophysics-guided approach combining docking, molecular dynamics simulations, and a series of assays with a particular emphasis on biophysical experiments. Several of the identified compounds induce the formation of aberrant capsids and inhibit HBV DNA replication in vitro, suggesting that they possess modest capsid assembly modulation effects. The synergistic computational and experimental approaches provided key insights that facilitated the identification of compounds with promising activities. The discovery of preclinical CAMs presents opportunities for subsequent optimization efforts, thereby opening new avenues for HBV inhibition.

RECOVER identifies synergistic drug combinations in vitro through sequential model optimization.

For large libraries of small molecules, exhaustive combinatorial chemical screens become infeasible to perform when considering a range of disease models, assay conditions, and dose ranges. Deep learning models have achieved state-of-the-art results in silico for the prediction of synergy scores. However, databases of drug combinations are biased toward synergistic agents and results do not generalize out of distribution. During 5 rounds of experimentation, we employ sequential model optimization with a deep learning model to select drug combinations increasingly enriched for synergism and active against a cancer cell line-evaluating only ∼5% of the total search space. Moreover, we find that learned drug embeddings (using structural information) begin to reflect biological mechanisms. In silico benchmarking suggests search queries are ∼5-10× enriched for highly synergistic drug combinations by using sequential rounds of evaluation when compared with random selection or ∼3× when using a pretrained model.

RetroGNN: Fast Estimation of Synthesizability for Virtual Screening and De Novo Design by Learning from Slow Retrosynthesis Software.

De novo molecule design algorithms often result in chemically unfeasible or synthetically inaccessible molecules. A natural idea to mitigate this problem is to bias these algorithms toward more easily synthesizable molecules using a proxy score for synthetic accessibility. However, using currently available proxies can still result in highly unrealistic compounds. Here, we propose a novel approach, RetroGNN, to estimate synthesizability. First, we search for routes using synthesis planning software for a large number of random molecules. This information is then used to train a graph neural network to predict the outcome of the synthesis planner given the target molecule, in which the regression task can be used as a synthesizability scorer. We highlight how RetroGNN can be used in generative molecule-discovery pipelines together with other scoring functions. We evaluate our approach on several QSAR-based molecule design benchmarks, for which we find synthesizable molecules with state-of-the-art scores. Compared to the virtual screening of 5 million existing molecules from the ZINC database, using RetroGNNScore with a simple fragment-based de novo design algorithm finds molecules predicted to be more likely to possess the desired activity exponentially faster, while maintaining good druglike properties and being easier to synthesize. Importantly, our deep neural network can successfully filter out hard to synthesize molecules while achieving a 105 times speedup over using retrosynthesis planning software.

The Mechanism of Action of Hepatitis B Virus Capsid Assembly Modulators Can Be Predicted from Binding to Early Assembly Intermediates.

Interfering with the self-assembly of virus nucleocapsids is a promising approach for the development of novel antiviral agents. Applied to hepatitis B virus (HBV), this approach has led to several classes of capsid assembly modulators (CAMs) that target the virus by either accelerating nucleocapsid assembly or misdirecting it into noncapsid-like particles, thereby inhibiting the HBV replication cycle. Here, we have assessed the structures of early nucleocapsid assembly intermediates, bound with and without CAMs, using molecular dynamics simulations. We find that distinct conformations of the intermediates are induced depending on whether the bound CAM accelerates or misdirects assembly. Specifically, the assembly intermediates with bound misdirecting CAMs appear to be flattened relative to those with bound accelerators. Finally, the potency of CAMs within the same class was studied. We find that an increased number of contacts with the capsid protein and favorable binding energies inferred from free energy perturbation calculations are indicative of increased potency.