Abstract. Raspberry ketone (RK) is a high-value flavor and fragrance compound widely used in the food, beverage, and cosmetic industries. However, existing

A sequence-based method dubbed AroMIP predicts the membrane insertion propensities of aromatic-centered motifs, thereby advancing mechanistic understanding of membrane insertion and enriching the tool...

Here, the authors develop a mouse model of wound infection to evaluate the efficacy of phage therapy for established infections and generate “HydroPhage”, a phage-containing hydrogel, as a means for s...

Effective and safe strategies to restore gene expression in haploinsufficiency disorders or to augment protective protein signaling have long been attractive but challenging therapeutic goals. Here, w...

Nature Reviews Bioengineering, Published online: 17 September 2026; doi:10.1038/s44222-026-00498-4Global access to diagnostics is constrained not only by the availability of testing technologies but also by their dependence on costly supply chains that typically originate from a small number of capital-intensive biotechnology hubs. Here, we argue that integrating decentralized diagnostics with local manufacturing and research capacity is essential to building resilient, equitable and sustainable diagnostic ecosystems.

Nature Reviews Bioengineering, Published online: 14 September 2026; doi:10.1038/s44222-026-00493-9Liquid–liquid phase separation (LLPS) is emerging as a key mechanism in cancer, linking biomolecular condensates to oncogenic signalling and tumorigenesis. This Review discusses evidence for dysregulated LLPS and highlights bioengineering strategies that enable therapeutic targeting of previously undruggable cancer drivers.

Aberrant activation of Wnt signaling results in unregulated accumulation of cytosolic β-catenin, which subsequently enters the nucleus and promotes transcription of genes that contribute to cellular p...

Excipients are widely used to suppress the self-association of therapeutic proteins, yet their mechanisms of action are not well understood and often assumed to be nonspecific. Here we show that ex...

In this Review, Iredell and colleagues examine key aspects of phage biology and their relevance to phage therapy, and outline current approaches for clinical application of phages, along with their as...

Microsystems & Nanoengineering - A smartphone-powered decentralized microfluidic platform for sensitive CRISPR-based nucleic acid detection

This Article demonstrates that direct preference optimization (DPO) can be used to effectively align an unsupervised structure-conditioned language model with biophysical information. The aligned mode...

Biological AI learns primarily from existing observations, but many questions cannot be answered from available data alone. Here we show that AI can instead acquire knowledge by acting directly on biological systems and learning from the consequences. We developed a closed-loop framework in which autonomous agents design protein variants, construct and characterize them in a robotic laboratory, learn from the resulting experimental feedback, and decide what experiments to perform next. We then allowed the system to operate continuously and without human intervention for approximately one month, during which multiple agents independently explored protein sequence space while learning from shared experimental experience. Applied to glycoside hydrolases, the agents discovered enzymes with substantially altered substrate specificity toward non-native sugars and progressively learned the structure of the underlying sequence-function landscape. The resulting experimental experience also revealed determinants of substrate specificity and protein expression that were not specified as learning objectives. These results demonstrate that AI can autonomously interact with biology over extended periods to acquire knowledge through experience, establishing a framework for biological discovery driven by continuous experimental interaction. ### Competing Interest Statement The authors have declared no competing interest. National Institute of General Medical Sciences, 5R01GM150929

DNA libraries are synthesized at scale.

Primary human myeloid cells are efficiently edited using virus-like particles.

An optimized strategy for the ribosomal synthesis of thioisoindole-bridged bicyclic (TiB) peptides is incorporated into the RaPID platform, enabling high-throughput discovery from topologically defin....

Synthetic macrocycles are promising therapeutics; however, most high-throughput discovery platforms rely on genetically encoded libraries of large peptide macrocycles, which are typically not optimize...

A membrane-anchoring nanodevice that integrates reversible, fuel-driven mechanical actuation enables programmable membrane penetration, cargo delivery and spatiotemporal regulation of biochemical reac...

Extensive conformational sampling identifies a folding-promoting disulfide bond that controls the oxidative folding pathway of disulfide-rich peptides. Exploiting this principle enables a generalizab....

RNA-based sensors offer powerful and programmable control of gene expression, yet our understanding of the structural principles that govern their potential design space remains incomplete. Here, we c...

Fluorescent proteins (FPs) are widely used reporters for visualizing cellular structures and processes. Traditional wet-lab strategies for FP engineering (rational design and directed evolution) have ...

Dogra et al. use super-resolution microscopy to uncover the molecular origins of spatial heterogeneity in the nucleolar granular component (GC). They show that differential interactions of nucleolar p...

Combining high-throughput molecular simulations, machine learning, and mixed-integer linear programming, the authors design peptides that localize to condensate interfaces, revealing surfactant-like, ...

Glycosylation is a structurally diverse, non-template-driven modification whose analysis by liquid chromatography-mass spectrometry is constrained by discovery-mode acquisition rules developed for pro...

Developing therapies and vaccines against integral membrane proteins is hindered by their extensive hydrophobic surfaces, which complicate production and structural analysis. Here, we describe a gener...

A suite of tools to improve translation and immunogenicity of circular RNAs, including nucleoside modification and rolling circle translation, for applications in cancer vaccines, protein replacement ...

Cell-free is a powerful platform to unravel biological complexity. Here, the authors used active learning to optimize buffer composition, boosting mRNA 20-fold, enabling cell-free transcriptomics and ...

Viral diseases continue to impose a major global health burden, yet substantial inequities in access to accurate molecular diagnostics persist, limiting effective surveillance and response. This Revie...

The manufacturing of cysteine-based antibody–drug conjugates (ADCs) requires precise control of the drug-to-antibody ratio (DAR), a critical quality attribute influencing product safety and efficacy. Conventional analytical techniques used for DAR determination, such as hydrophobic interaction chromatography (HIC) and reversed-phase high-performance liquid chromatography (RP-HPLC), are low-throughput and unsuitable for real-time process monitoring, limiting process development efficiency. To address this challenge, we developed a high-throughput size-exclusion chromatography–mass spectrometry (SEC-MS) workflow with and without postcolumn denaturation (PCD) functionality, enabling real-time monitoring of both reduction and conjugation kinetics during cysteine-based ADC manufacturing. Using a short SEC column coupled to a mass spectrometer, analysis time can be performed under 2 min per sample. This workflow was applied to rapidly optimize critical process parameters, including reductant concentration and pH during reduction, and drug-linker ratio during conjugation. The approach provided deeper mechanistic insights into reaction kinetics and supports process analytical technology (PAT)-enabled strategies for advanced process control. Overall, this high-throughput SEC-MS workflow accelerates ADC development while improving manufacturing robustness.

Jumbo phages are bacterial viruses with double-stranded DNA genomes exceeding 200 kb. In this Review, Sam van Beljouw & Li Yuping examine the biology of these bacteriophages, highlighting their di...

Abstract. Riboswitches are useful models for revealing how some RNA molecules undergo dynamic rearrangements of their structures to perform cellular functi

Bacteriophage receptor-binding proteins (RBPs) determine bacterial host recognition and are central to phage host-range evolution, yet the structural principles governing how RBPs diversify and expand...

There have been a few recent works showing the advantages of combining cell-free protein expression (CFE) with biolayer interferometry (BLI) for the rapid kinetic characterization of novel capture pro...

Protein binder design has largely optimized for affinity alone, leaving conformational selectivity unaddressed: for allosteric targets such as kinases, nuclear receptors, and GPCRs, a binder that enga...

Therapeutic mRNA design requires coordinating multiple interacting sequence features across the full transcript, where codon usage, untranslated regions (UTRs), and their coupling jointly determine st...

The success of mRNA-lipid nanoparticles (LNPs) vaccines during the COVID-19 pandemic has significantly increased global demand for stable, thermo-resi…