Lead optimization articles from across Nature Portfolio

Bivalent molecules comprising dual functional motifs often exhibit strong target binding, however arriving at the optimal structure is challenging. Here, the authors employ EGFR as a model system for understanding a subtle difference in the linker structure of bivalent inhibitors to enhance binding against drug-resistant EGFR mutants.

Human interleukin-10 (IL-10) is known to be an immunosuppressive cytokine for anti-inflammation, however, IL-10 can enable the evasion of tumor cells through immunosuppression of the host immune system. Here, the authors develop an IL-10 inhibitory peptide, NK20a, that displays anticancer effects, and show that its conjugation with chemotherapeutic gold nanoparticles enhances their anticancer efficacy.

Late-stage functionalization of complex drug molecules is challenging. To address this problem, a discovery platform based on geometric deep learning and high-throughput experimentation was developed. The computational model predicts binary reaction outcome, reaction yield and regioselectivity with low error margins, enabling the functionalization of complex molecules without de novo synthesis.

Late-stage functionalization of drug molecules can tune their properties without the need for entirely new syntheses, however, predicting reactivity and planning synthesis for late-stage C-H activation remains challenging. Here, the authors develop a reaction screening approach combining high-throughput experimentation with computational graph neural networks to identify suitable substrates that can be used for late-stage C-H alkylation via Minisci-type chemistry.

A physics-informed deep learning model, PBCNet, is proposed for predicting the relative binding affinity of ligands in order to improve guiding structure-based drug lead optimization.

A pairwise binding comparison network (PBCNet) has been established for predicting the relative binding affinity among congeneric ligands, using a physics-informed graph attention mechanism with a pair of protein pocket-ligand complexes as input. PBCNet shows practical value in guiding structure-based drug lead optimization with speed, precision, and ease-of-use.

The therapeutic applications of DNAzymes are limited because of their low effectiveness in vivo. Now, a promising approach for constructing DNAzymes that show high gene-silencing efficiency in mammalian cells has been developed. This approach incorporates chemical modifications into an existing DNAzyme scaffold.

A new click-style reaction based on a strain-release amination strategy has been developed. This approach can be used to append small, strained ring systems onto a core scaffold.