South Korea has positioned hydrogen and carbon capture, utilization and storage (CCUS) as key technological pillars of its 2050 carbon neutrality strategy. Despite sustained and ambitious policy intervention, however, empirical evidence on how policy interventions have reshaped public R&D portfolios and early industrial diffusion remains limited. This study examines hydrogen and CCUS in South Korea through a policy−R&D−industrial diffusion framework, focusing on policy sequencing and institutional design. The analysis combines qualitative content analysis with project-level R&D data from the National Science and Technology Information Service (NTIS) and interrupted time series analysis to identify policy-induced structural breaks, while industrial diffusion is assessed using deployment indicators such as hydrogen vehicles, hydrogen refueling stations, and large-scale (>100,000 tons/year) CCUS projects. The results indicate a clear divergence in policy outcomes between the two sectors. Hydrogen policy interventions rapidly shifted public R&D toward utilization and industry-oriented development stages and were accompanied by a sharp increase in public R&D expenditures, thereby accelerating deployment. By contrast, CCUS experienced fragmented R&D adjustments, continued public-sector dominance, and limited large-scale demonstration. These findings highlight that while policy-driven, mission-oriented innovation pathways can mobilize R&D and early markets, their effectiveness depends on timely legal frameworks, the institutional coherence of the policy framework, and market-supporting infrastructure. In the case of CCUS, delayed legal institutionalization has constrained the transition from public R&D accumulation to industrial-scale deployment, highlighting the need for stronger regulatory clarity, market-creation instruments, and supporting data infrastructures. • Green innovation follows a policy-driven, top-down pathway in South Korea. • Study links policy, R&D, and industry using legislative and project-level data. • Hydrogen policy synchronization accelerated market-oriented R&D and deployment. • Delayed CCUS legislation caused policy lags and fragmented R&D trajectories. • Innovation requires timely legal frameworks and supportive market environments.

Deep learning unlocks sequence-divergent synthetic promoters to empower Streptomyces natural product engineering.

Extrahepatic delivery of small interfering RNA (siRNA) remains a major translational challenge because most nanocarriers preferentially accumulate in the liver, while endosomal sequestration limits productive cytosolic release. Inflammatory macrophages in the spleen are attractive therapeutic targets in systemic inflammation, yet spleen-selective delivery systems with efficient endosomal escape remain underdeveloped. Here, a structure-guided peptide engineering workflow was used to generate histidine-rich, pH-switchable endosomolytic peptides for spleen-selective siRNA delivery. Sequence design integrated pH-dependent charge transition modeling, amphipathic helix prediction, membrane interaction scoring, and safety filtering. Six candidate peptides were synthesized and evaluated for pH-responsive structure, membrane disruption, hemocompatibility, siRNA complexation, serum stability, macrophage uptake, endosomal escape, biodistribution, and anti-inflammatory efficacy. The lead peptide, HSEP-6, showed a predicted net charge increase from +3.1 at pH 7.4 to +7.4 at pH 5.5, helix content increasing from 17% to 56%, and acidic calcein release increasing from 9% to 62%. In inflammatory macrophages and LPS-challenged mice, HSEP-6 enabled efficient siRNA delivery, spleen-selective accumulation, marked Irf5 silencing, reduced TNF-α and IL-6, and no measurable systemic toxicity, supporting histidine-rich pH-switchable peptides as a rational platform for extrahepatic RNA delivery.

ATP-responsive injectable DNA hybrid hydrogels for regulating the "on-demand release" of fibroblast growth factor 21 cytokines to facilitate muscle regeneration.

Molecular property prediction is essential in drug discovery for early-stage compound evaluation. Recently, contrastive learning has demonstrated significant potential under limited labeled data by constructing augmented views. However, current augmentation strategies often disrupt molecular semantics and ignore chemical priors, limiting representation quality. Moreover, molecular data is inherently multimodal, including graphs, fingerprints, and sequences, yet how to effectively integrate their complementary information remains challenging. Therefore, we propose MPMFMol, a unified framework that integrates multitask self-supervised pretraining with multimodal fine-tuning for molecular property prediction. During pretraining, we construct heterogeneous augmented views based on molecular fragments to preserve original molecular semantics, enabling the graph encoder to capture fragment-level information. Meanwhile, fingerprint features are integrated into a multitask learning objective, reducing reliance on negative sampling and enhancing the encoder's representation capability. During fine-tuning, we further incorporate functional group and SMILES sequence information and design a stage-aware modality fusion strategy. Specifically, pretrained graph features are injected into the initial representation of functional groups to guide feature extraction and then fused with SMILES features to enable deep cross-modal interaction and enhance downstream predictive performance. Experimental results on six classification and three regression data sets demonstrate that MPMFMol outperforms state-of-the-art baselines.

To provide explicit, vendor-independent view-order constructions that enable selection of an arbitrary center echo in multi-shot imaging, with a focus on RARE and MPRAGE. Vendor-independent view-order methods were developed to assign echo and shot indices to a prescribed set of phase encodes, enabling explicit selection of the center echo and avoiding large discontinuities near k-space center. The view-order methods were implemented in 2D RARE, 3D RARE, and MPRAGE, and evaluated in phantom and in vivo experiments, where the effects of the different methods were explored. An interactive dashboard was provided to reproduce them. All methods enabled placing the chosen echo/shot index at the k-space center without changing other scan parameters. In RARE, varying the center echo produced the expected shift in weighting. In MPRAGE, varying produced the expected shift in weighting, in agreement with simulations. Artifact patterns in RARE were consistent with discontinuities and symmetry properties of the echo-index distribution. Arbitrary-center view orders decouple scan parameters such as echo train length, matrix size, and undersampling factors, from center-echo selection. The presented constructions are compatible with common sampling strategies (e.g., partial Fourier, parallel imaging, and elliptical coverage) and sequences. Arbitrary center echoes allow for more control over image contrast and present an additional degree of freedom in sequence design.

Chiral ensembles can arise through supramolecular curvature that resolves geometric frustrations in the packing of bent, achiral molecular or colloidal building blocks. Here, we leverage orthogonal protection-deprotection click chemistry to create sequence-defined mesogenic heterodimers exhibiting emergent chirality. We compare the hierarchical self-assembly of the synthesized asymmetric, achiral heterodimers, which differ only in the position of a methyl substituent. Both dimers form chiral spherulites composed of nanocylinders. However, the detailed arrangement of nanocylinders depends on the position of the methyl substituent and the crystallization conditions. Despite the chemical similarity, in one dimer, two crystalline forms are optically active. They form conglomerates of dextrorotatory and levorotatory spherulites. The other dimer forms more highly anisotropic spherulites that mask circular birefringence arising from the misorientation of nanocylinders, while mapping of nanocylinder directors reveals a sense at the spherulite surface. We propose that differences in nanocylinder arrangements may arise from changes in nanocylinder curvature and dimensions dictated by the methyl substituent position, inducing chirality. These results demonstrate multiscale hierarchical assembly relevant to dense systems of tubular structures and highlight the role of sequence and molecular design in directing the bottom-up hierarchical self-assembly and chirality of mesogenic systems.

Heparan sulfate (HS) proteoglycans are abundant, sulfation-patterned glycosaminoglycans on brain cells and the neurovascular unit that interact with amyloid-β (Aβ) and tau, aggregate and propagate proteopathic species, and modulate cellular uptake pathways. Aptamers are short, chemically synthesized single-stranded nucleic acids that fold to form high-affinity ligands. Novel aptamers bind specific HS motifs for targeting ligands to deliver RNA therapeutics (siRNA, antisense oligonucleotides, mRNA, RNA aptamer-cargo hybrids) into the brain and diseased cells in Alzheimer's disease (AD). We cover diagnostic aptamers, SELEX approaches adapted for glycans and HS, chemistry, optimization (AI/machine-learning-enabled design), novel therapeutic applications, delivery vehicles and conjugation strategies, catalytic aptamers/ribozymes, and challenges for clinical translation. We propose a workflow-in-silico HS motif mapping and machine learning (ML)-guided sequence design through microfluidic/in-vivo SELEX, site-specific chemical modification for nuclease resistance and blood-brain barrier (BBB) transcytosis, to scalable GMP manufacture. Aptamer-HS technology could enable the precision delivery of RNA therapeutics in AD.

DNA walkers can drive nucleic acids to move along designed tracks or processes in DNA circuits, endowing the circuits with highly efficient signal-amplification capability. However, their actuation and movement are difficult to regulate, and integrating them into circuits for high-throughput detection remains challenging. In this work, we propose a novel DNA nanomachine termed "DNA stalker", which can both actuate and regulate the movement of DNA walker while conferring multiplex detection capability to the entire circuit. Herein, DNA stalker can regulate DNA walker via target-responsive catalytic hairpin assembly, and thereafter be integrated with MXene to fabricate the MXene-supported DNA circuits. These DNA circuits operate in an enzyme-free manner, and the inherent limitations of current DNA walkers are overcome solely through rational nucleic acid sequence design. In addition, program-customized design enables the tailored fabrication of the MXene-supported DNA circuits according to the customer-specified target combinations to be detected, without requiring prior knowledge of nucleic acid sequence design when the detection targets are switching. To meet the requirement of clinic detection, a microfluidic device is employed to operate three circuit sets with independent outputs simultaneously, enabling the concurrent detection of six miRNAs in serum samples from healthy individuals and patients with three different types of cancer (breast, cervical and lung cancer). This system achieves highly sensitive and high-throughput detection of miRNAs in serum, providing a powerful tool for the early clinical diagnosis of cancer at the molecular level.

ClinASO: An open-source platform for rapid drug discovery of gapmer antisense oligonucleotides.

Aldo-keto reductase AKR13B3 catalyzes the reduction of 3-keto-deoxynivalenol (3-keto-DON) to nontoxic 3-epi-DON, thereby alleviating mycotoxin contamination in food and feed. However, the previously reported AKR13B3 mutant AKR-V3 exhibited high catalytic activity but poor thermostability, restricting its industrial application. Here, guided by molecular dynamics flexibility analysis and consensus sequence design, we engineered the variant M5 with markedly improved thermostability. M5 showed a 38.3-fold longer half-life at 60 °C and retained slightly enhanced catalytic efficiency. Quantum mechanics/molecular mechanics (QM/MM) calculations revealed that M5 possessed a lower activation barrier than AKR-V3, which explained its superior activity retention under high-temperature conditions. Moreover, structural analyses and molecular dynamics (MD) simulations further demonstrated that the enhanced thermostability of M5 stemmed from reinforced hydrogen bonding, increased hydrophobic packing, elevated structural rigidity, and optimized surface electrostatic potential. This work provides an effective strategy for designing thermostable detoxification enzymes to eliminate mycotoxins in industrial applications.

G-triplexes are noncanonical DNA structures formed by three guanine tracts, but their sequence design rules for small-molecule binding remain unclear. In this work, we carried out a systematic sequence screening to optimize G-triplex sequences for improved interaction with methylene blue (MB), and applied the best sequence in an electrochemical biosensor for melamine detection. Over 80 candidate sequences were designed by varying loop nucleotides, 5'-flanking bases, and strand length, and were evaluated by square wave voltammetry. The optimal sequence F9 (5'-ATGGGAGGGTGGG-3') achieved the highest current suppression (ΔI/I0 ≈ 0.81), outperforming all tested G-triplex and several well-known G-quadruplex sequences. Circular dichroism confirmed that F9 folds into a parallel G-triplex with a melting temperature of 65.1 °C, which increased to 70.0 °C upon MB binding. Electrochemical titration gave a 1:1 binding ratio with an association constant of 9.37 × 105 M-1. Molecular dynamics simulations showed that MB binds to the 5'-face of the G-triplex by π-π stacking, with a computed binding free energy of -151.9 kJ/mol. A hairpin probe (MelaPin) combining the F9 sequence with a poly-T melamine recognition region was then constructed. The resulting biosensor showed a linear response over 1-100 μM melamine with a detection limit of 0.74 μM, and a recovery of 96.0 ± 0.8% in spiked infant formula. This work provides a practical framework for G-triplex sequence design and shows that the G-triplex/MB system can serve as a simple, label-free signal module for electrochemical sensing.

Generative AI algorithms such as the transformer and diffusion models have greatly empowered de novo design of proteins capable of specifically interacting with designated structural sites on another protein. Most of these design methods employ a top-down approach, in which an overall protein shape is generated by an AI model to pack against a given structural site, followed by sequence design to optimize the interaction. Despite being trained on limited protein complex structures available in the database, the top-down approach has yielded encouraging results. Here, we propose a bottom-up approach that generates atom clusters for optimal packing against a specified structured region for informing the design of protein-protein interactions. To this end, we trained a masked discrete diffusion model, named Void-X, that uses the diffusion transformer to learn atomic-level interactions and fill atomic voids in protein interaction interfaces. Void-X was trained using 8.7 million spherical clusters of atoms from experimental structures in the Protein Data Bank. In each cluster, ~70% of the atoms are used as context (or prompt), and ~30% are masked for information recovery (or answer). By training the model with 172 million parameters, Void-X achieves an overall accuracy of 78.3% and 68.2% for intra- and interchain spherical clusters, respectively. Furthermore, we find that information entropy is a reliable indicator of the prediction accuracy for Void-X. This level of performance allows de novo generation of molecular interactions at the atomic level, offering an alternative approach of protein design complementary to the existing ones.

Transforming mRNA drug design with AI: From UTR and codon optimization to coordinated design.

Native homo-oligomeric transmembrane β-barrels (TMBs) have been widely explored for molecular sensing, sequencing, and separation applications, but their uniform lumen limits spatial resolution and the localized interactions required for both analyte discrimination and filtration. Monomeric TMBs have been designed using energy-based methods but this has required extensive expert input, limiting scalability and functionality. Here, we present a rapid generative AI method for TMB design that combines diffusion-based backbone generation conditioned on β-barrel structural features with TMB-optimized sequence design. We characterized 48 designs (spanning 10-16 strands) that exhibited measurable conductances pertaining to pore diameters of 0.7-1.5 nm, and determined crystal structures of two designs that have atomic-level agreement with the designed models. We demonstrate the versatility of the generative method by designing nanopores with copper binding sites for selective ion sensing, larger pores that support DNA translocation, and longer pores with increased hydrophobic thickness, that mediate ion transport across three-dimensional networks of monoglyceride bilayers and synthetic polymer-lipid hybrid membranes composed of block copolymers.

Despite more than half a century of research on selenoproteins, the central catalytic intermediate, selenocysteine selenenic acid (Sec-SeOH), has remained experimentally elusive. Its isolation has long been impeded by its presumed instability and propensity for thermal deselenation. Here, we report the first isolable Sec-SeOH at ambient temperature. This relies on a bioinspired design of a selenopeptide sequence encapsulated within a protective cradle, together with an oxidant-free route from the corresponding selenenyl iodides (Sec-SeI), enabling x-ray structural analysis and chemical characterization. The isolated Sec-SeOH shows unexpected resistance to β-elimination to dehydroalanine (DHA). Oxidation experiments combined with theoretical calculations demonstrate that conversion to DHA proceeds preferentially via overoxidation to the seleninic acid (Sec-SeO2H), for which β-elimination is substantially more favorable. Reactivity profiling further highlights the pronounced electrophilicity of Sec-SeOH toward biologically and pharmacologically relevant nucleophiles. These findings redefine the stability-reactivity landscape of Sec-SeOH and provide a foundation for strategies aimed at preventing selenoprotein inactivation. Beyond defining an isolable Sec-SeOH model, the work provides a molecular-level rationale for how selenoproteins can combine high selenium-centered reactivity with resistance to irreversible oxidative self-inactivation.

Discovery of TCR-like antibodies to the KRAS G12D neoantigen via in silico-in vitro workflow.

Zero echo time magnetic resonance imaging is an ultrashort echo time technique that enables computed tomography-like visualization of cortical and trabecular bone while preserving the ionizing radiation-free, multiplanar, and soft tissue advantages of conventional magnetic resonance imaging. This narrative review outlines the technical foundations of spinal zero echo time imaging, such as sequence design, key optimization parameters, advanced deep learning-based reconstruction strategies, and common interpretive pitfalls like gas mimicking calcification, metal-related artifacts, and limited spatial resolution. Clinical applications are detailed across traumatic, degenerative, inflammatory, neoplastic, infectious, and developmental spinal conditions, emphasizing how zero echo time complements standard magnetic resonance imaging sequences and often approaches computed tomography in depicting cortical disruption, osteophytes, erosions, sclerosis, and heterotopic ossification. I also highlight emerging roles for zero echo time in pediatric and fetal spine assessment and in magnetic resonance imaging-only surgical planning workflows, where accurate computed tomography-like bone contrast is required but ionizing radiation avoidance is paramount. Collectively, current evidence positions zero echo time as a versatile adjunct that can streamline imaging algorithms, reduce dependence on computed tomography, and support ionizing radiation-sparing magnetic resonance imaging focused pathways for diagnosis, treatment planning, and follow-up in spine disease.

DNA nanostructures have emerged as a highly programmable platform with applications spanning nanomedicine, biosensing, and molecular machinery. Although their design is primarily governed by sequence-encoded base pairing, growing evidence indicates that the chemical environment critically modulates their assembly, stability, and function. This review systematically summarizes how key environmental factors, including pH, cation identity, and additives, influence DNA nanostructures across multiple hierarchical levels. We begin by outlining the fundamental principles of DNA assembly, with emphasis on sequence design, sticky end interactions, and DNA concentration. We then discuss the effects of pH, cation identity, and additives on structural stability and dynamic behavior. Building on these insights, we further explore how environmental responsiveness is harnessed to construct advanced systems for DNA-based nanodevices, delivery, biosensing, and environmental monitoring. Finally, current challenges and future perspectives are outlined. This review provides a comprehensive framework for understanding and exploiting chemical environment-mediated regulation in DNA nanotechnology, paving the way for the rational design of adaptive and functional nanomaterials.

With the emergence of powerful deep learning-based tools, computational protein design has become a widely accessible technique. Nowadays, it is possible to perform both sequence and structure design in a matter of minutes, making the technology attractive to the broader scientific community. In protein design campaigns, one of the most common in silico strategies to evaluate how well a sequence encodes a target structure is the so-called self-consistency or refolding pipeline. In this approach, a structure prediction model is used to refold the designed sequence to probe whether it is compatible with the intended structure, and is evaluated via two metrics linked to experimental success: the confidence score of the predicted structure (predicted local distance difference test) and the self-consistency root-mean-square deviation, which measures how closely the refolded structure matches the target. In this work, we systematically evaluate how different models and structure prediction settings impact these metrics, and to what extent they can be used to reliably filter sequence design candidates. We show that evolutionary information can obscure folding models' abilities to assess sequence-structure compatibility, reducing the predictive performance of refolding metrics for experimental success, particularly for designs that share homology with natural sequences. We further highlight limitations of refolding metrics, including their sensitivity to structural features, such as flexibility. Our findings raise awareness of potential pitfalls in refolding-based evaluation and support more informed use of these metrics in protein design campaigns.