Biomanufacturing is undergoing a profound paradigm shift from traditional experience-driven methods to innovative data-driven approaches,with artificial intelligence(AI)serving as a central enabling force.This article systematically reviews breakthroughs and applications of AI across key domains of biomanufacturing.In biological component design,generative AI facilitates"from-scratch"innovation for creating DNA regulatory elements,signal peptides,and enzymes,while continuously refining functionality through intelligent"design-build-test-learn"closed-loop systems.This shift enables researchers to move beyond imitation of natural systems toward truly rational and programmable biological design.In metabolic network reconstruction,multi-scale metabolic models combined with advanced deep learning frameworks like AlphaGEM enable dynamic simulation of complex metabolic behaviors,cross-species generalization,and exploration of previously uncharted biological functions.These computational advancements are particularly significant for engineering non-model organisms with industrial relevance,where traditional characterization methods are often inadequate.Within cell-free synthesis systems,AI converges with microfluidic and automation platforms to establish high-throughput,intelligent,and self-optimizing protein production pipelines,effectively decoupling biological synthesis from the constraints of cellular viability and regulation.This approach enables rapid prototyping of biomolecules that would be difficult or impossible to produce using conventional cellular systems.For process optimization and control,AI drives the evolution toward rationally designed,dynamically optimized,and autonomously regulated production through predictive modeling,real-time sensor integration,and adaptive control strategies.These systems continuously learn from process data to maximize yield,minimize variability,and ensure consistent product quality while reducing resource consumption and operational costs.While AI has substantially accelerated biomanufacturing research and development(R&D)and enhanced production efficiency,several challenges remain,including incomplete understanding of underlying biological mechanisms,scarcity of high-quality annotated datasets,limited model interpretability(often described as the"black box"problem),complexities in multi-objective optimization,and stringent real-time operational demands in industrial settings.Looking forward,deeper integration of AI with synthetic biology,robotic automation,and multi-omics data analytics will propel biomanufacturing toward a programmable,intelligent,and sustainable next-generation industrial ecosystem.Future advancements are expected to include more sophisticated physics-informed neural networks that incorporate fundamental biological principles,federated learning approaches to leverage distributed data while maintaining privacy,and the development of digital twins for entire bioprocesses.This convergence is poised to deliver a key technological engine for the green transformation of critical sectors such as pharmaceuticals,chemicals,and energy.Ultimately,the synergy between AI and biotechnology promises to establish a more efficient,resilient,and environmentally conscious bioeconomy capable of addressing global challenges in health,materials,and sustainable production.
Brain-computer interface(BCI)represents a strategic frontier technology driving a new wave of scientific and technological revolution and industrial transformation.In China,BCI has been designated as one of the six industries of the future.The year 2025 marks a historic transition of BCI technologies from laboratory-based exploration to large-scale clinical deployment,with the clinical value of BCI systematically validated across multiple domains and scenarios.The field has formally entered into a new phase of practical development driven by clinical needs.This article analyzes the innovation ecosystem of BCI,highlights major advances achieved in 2025,and explores future development trajectories.The BCI innovation ecosystem comprises three layers.The foundational layer encompasses two pillars:brain science and machine-related technologies.The core technology layer covers the technical pipeline from signal acquisition and processing to execution and feedback.The application layer is centered on healthcare while expanding into diversified domains,including education and consumer applications,industrial manufacturing,intelligent transportation,and military and security.Major advances in 2025 are reflected in several aspects.First,systematic progress has been achieved in brain atlases and perceptual-cognitive mechanisms,providing a theoretical basis for BCI.Second,in core technological innovation,multimodal signal integration has accelerated;hardware such as electrodes and chips has rapidly evolved toward minimally invasive,highly integrated,low-power,and high-throughput designs;and large brain models have driven signal decoding toward higher accuracy,stronger generalizability,improved robustness,and scenario-oriented deployment.Third,clinically,BCIs have advanced from restoring motor,language,and visual functions to enabling precise closed-loop modulation of neuropsychiatric disorders,while also achieving scaled applications in consumer,industrial,educational,and national security contexts.Fourth,policies,standards,reimbursement systems,capital investment,and platform infrastructures have jointly established a robust innovation foundation,fostering a relatively mature industrial ecosystem.Looking ahead,diverse technological pathways are expected to advance in parallel,with deep integration of biological intelligence and artificial intelligence giving rise to a brain-machine intelligence symbiosis.Application scenarios are anticipated to expand rapidly,evolving from"functional substitution"to"capability enhancement"and ultimately to"experience creation".Industrial competition will increasingly take the form of competition among ecosystems.In this rapidly evolving landscape,data and algorithms alongside devices will emerge as core assets,accompanied by complex ethical and legal challenges that necessitate the development of comprehensive governance frameworks.
Virtual clinical trials(VCTs),as a computational model-centered tool,are progressively transforming the traditional research and development(R&D)paradigms for drugs and medical devices.With the iterative advancement and integrated application of computer modeling and simulation technologies,artificial intelligence(AI),and digital twin(DT),the VCT technology framework is continuously being enriched and refined.These diversified technologies continuously empower the entire lifecycle R&D of drugs and medical devices,with applications expanding and deepening across stages including drug discovery,preclinical research,and clinical trials,significantly enhancing R&D efficiency and reducing costs.The regulatory framework has been further refined,standards such as model credibility assessment are being gradually established,and the application of computer models and AI-generated virtual evidence in product evaluation and approval is receiving increasing attention.Despite existing challenges related to data foundations,model interpretability,regulatory standards,and technology translation,future advancements will rely on the integrated verification of multi-scale models,the construction of standardized data ecosystems,and the coordinated promotion of global regulatory harmonization.VCTs are expected to profoundly empower the transformation and upgrading of biomedical R&D.
The convergence of artificial intelligence(AI)and life sciences has evolved from unidirectional tool-assisted research into a bidirectional,co-evolutionary paradigm characterized by mutual knowledge transfer and synergistic advancement.This review aims to systematically examine the latest progress in this interdisciplinary fusion,delineate the underlying logic of their interactive empowerment,and discuss the transformative potential and remaining challenges of this symbiotic relationship for future scientific research.The foundational logic of AI-life science integration rests on three pillars:the analogy between biological and computational information-processing paradigms,a shared mathematical language for modeling systemic complexity,and a spiraling cycle of bidirectional knowledge transfer from"AI understanding life"to"life inspiring AI".Building on this foundation,the review first surveys the innovative development of Generative Biology across multiple scales.At the molecular level,breakthroughs such as AlphaFold-series models and diffusion-based protein design frameworks(e.g.,RFdiffusion)have fundamentally shifted the paradigm from observation to de novo creation,enabling atomic-precision design of proteins,enzymes,and drug candidates.At the cellular level,single-cell foundation models and virtual cell frameworks leverage deep generative architectures to learn transferable cell-state representations,supporting perturbation response prediction and causal mechanistic reasoning.At the organ level,deep learning empowers medical imaging analysis,organ-on-chip integration,and digital twin construction for individualized intervention simulation.At the individual and ecological levels,AI-driven models integrate electronic health records,wearable sensor streams,wastewater-based epidemiological surveillance,and biodiversity monitoring to bridge data gaps across personal health,public health,and environmental stewardship.The review then examines Bio-inspired AI,encompassing brain-inspired learning rules and neuromorphic hardware that replicate neural computation with event-driven,energy-efficient architectures,biomimetic optimization algorithms including genetic algorithms,artificial immune systems,and swarm intelligence,as well as molecular computing and biological hybrid intelligence platforms that employ DNA circuits and living organoid networks as computational substrates.Despite remarkable progress,critical challenges persist,including the"black-box"accountability dilemma in clinical and regulatory contexts,the corrosive risks of AI hallucinations in molecular and clinical design,algorithmic adaptability to the inherent fuzziness and nonlinearity of biological systems,and multi-scale data heterogeneity.Addressing these challenges demands the establishment of robust"dry-wet closed-loop"validation systems,transparent and explainable model architectures,and globally coordinated governance frameworks that balance innovation incentives with biosecurity,privacy protection,and algorithmic fairness,thereby ensuring a safe and sustainable trajectory for this transformative convergence.
The life sciences research paradigm is undergoing a profound transformation from unidirectional data analysis towards a synergistic,closed-loop system of"data-model-knowledge-data".This evolution is centrally driven by the pervasive integration of artificial intelligence technologies,which are redefining biological data from static repositories into programmable,designable intelligent entities.This paper systematically examines the bidirectional transformation between biological data and knowledge,highlighting the critical roles of AI-ready data,intelligent models,and the"Learn-Design-Build-Test"(LDBT)cycle.In the Data-to-Knowledge(D2K)trajectory,the journey begins with ensuring data"AI-ready",adhering to FAIR principles,possessing standardized formats,and being semantically aligned with biological knowledge.High-quality,structured data from major databases like PDB,NCBI,and GEO fuel sophisticated models.These models learn patterns to generate statistical or correlative knowledge.The crucial next step,Model-to-Knowledge(M2K),involves translating model outputs into verifiable scientific knowledge,such as mechanistic hypotheses.Enhanced model interpretability and integration into the LDBT cycle are essential for this transformation,moving beyond mere correlations to testable biological insights.Conversely,the Knowledge-to-Data(K2D)trajectory initiates with Knowledge-to-Model(K2M),where established mechanistic,associative,or hypothetical knowledge is encoded into computational model architectures.This is exemplified by digital twins and virtual cell models,which embed biological priors as structural constraints.Subsequently,in Model-to-Data(M2D),these knowledge-informed models including generative AI like diffusion models,cross-omics translators,and single-cell foundation models actively synthesize biologically plausible predictive or synthetic data.This addresses data scarcity and guides experimental design.The LDBT paradigm forms the core operational engine that unifies these bidirectional paths,creating a spiraling iterative relationship.Data drives model learning,models distill knowledge,and knowledge feeds back to generate new data for training superior models.However,challenges remain,including ensuring the reliability and reusability of AI-extracted knowledge,bridging the"conversion gap"between computational designs and successful experimental validation,and establishing standardized interfaces between D2K and K2D stages.Looking forward,the bidirectional loop is posited as a fundamental methodological framework for tackling biological complexity and integrating multimodal data.Its systematic engineering,through the continuous optimization of the LDBT cycle within research infrastructure,paves the way for life sciences to advance into an era of predictive and designable intelligence.Future efforts must focus on building a robust AI-ready data foundation,developing next-generation algorithms that deeply integrate data and prior knowledge,and perfecting the dry-wet lab integration for automated scientific discovery.
Driven by dual impetuses of supportive policies and cutting-edge technological breakthroughs,the global and Chinese radiopharmaceutical(nuclear medicine)sectors have undergone profound innovative leaps and transformative paradigm shifts in recent years.Moving beyond their traditional foothold in well-established oncology indications,radiopharmaceutical products are now expanding their utility into a diverse spectrum of solid tumors and a broad range of non-oncological therapeutic areas.This paper is dedicated to systematically sorting out and analyzing the pivotal advances and evolutionary trends shaping the global and Chinese radiopharmaceutical landscapes in 2025,against the backdrop of the deepening integration of diagnostic and therapeutic functionalities.These critical developments are manifested in three core dimensions.First,in terms of radionuclide supply,China has achieved a landmark breakthrough in the domestic mass production of lutetium-177(177Lu),a workhorse radionuclide for targeted therapies.In contrast,actinium-2 2 5(225Ac)—the core radionuclide underpinning next-generation targeted alpha-particle therapy—continues to face severe constraints stemming from global supply chain bottlenecks,yet it has already become a strategic focal point for multinational pharmaceutical corporations vying for competitive edges in the field.Second,regarding clinical translation,radiopharmaceutical conjugates(RDCs)have not only solidified their proven efficacy advantages in conventional tumor types but also pioneered innovative dual-targeting strategies tailored for refractory subtypes such as neuroendocrine differentiated prostate cancer(NEPC).Moreover,RDCs are demonstrating burgeoning exploratory potential in non-oncological domains,including cardiovascular disorders and neurodegenerative diseases.Third,on the industrial ecosystem front,China has initially constructed a comprehensive full-chain layout encompassing isotope production,drug research and development,and global trade cooperation,with its domestic market witnessing exceptionally robust growth rates.Nevertheless,the industry is confronted with pressing challenges,most notably the acute shortage of highly skilled professional talents,which demands urgent resolution.Grounded in the core development philosophy of"breaking bottlenecks,consolidating strengths,and addressing shortcomings",this paper puts forward targeted recommendations for future development:efforts should be concentrated on establishing a stable and resilient radionuclide supply system,expanding the clinical application scenarios of RDCs across multiple disease categories,intensifying the cultivation of interdisciplinary talents,and leveraging policy guidance and resource integration to accelerate the transition toward a new era of individualized precision stratified therapy,which is underpinned by the integration of multi-omics technologies.
In recent years,the convergence of biological and non-biological technologies has been reshaping the global landscape of scientific and technological development with unprecedented breadth and depth,emerging as a core driving force for fostering disruptive innovation,cultivating new quality productive forces,and advancing sustainable transformation.This integration transcends the simplistic combination of isolated technological functions typical of earlier stages,gradually evolving into a systemic,multi-scale collaborative innovation paradigm.It provides robust support for cross-level systematic research from molecules to individuals,ranging from precise regulation at molecular interfaces,intelligent construction of cellular microenvironments,and in vitro simulation of organ functions to digital mapping of life systems at the organismal level,thereby driving a shift in life science research from traditional"passive observation"toward"active design"and even"dynamic optimization".This article focuses on this cutting-edge trend.First,it delves into the underlying logic of the convergence between biological and non-biological technologies,highlighting that the core lies in establishing system-level coupling mechanisms that bridge living systems and artificial systems across physical,informational,and functional dimensions.Such coupling is not merely about interface connections;rather,it involves the efficient integration of signal transduction,energy exchange,and functional synergy at the bio-nonbio interface through multidisciplinary tools such as biomimetic design,synthetic biology,micro/nano-fabrication,and artificial intelligence.Second,the article examines several key nodes and representative application areas,systematically reviewing advances in frontier fields such as programmable biomaterials,biosensors,3D/4D bioprinting,organoids,brain-computer interfaces,biohybrid robots,and digital life.For instance,programmable biomaterials enable fine-tuned control over material structure and performance in response to external stimuli;biosensors facilitate real-time,highly sensitive monitoring of cellular metabolism or neural activity;3D/4D bioprinting significantly enhances the biomimetic fidelity and controllability of organoids at the tissue level and in terms of cellular heterogeneity;biohybrid robots demonstrate potential in precise manipulation,environmental detection,micro-scale operation,and life-like system modeling;and digital life technologies propel life sciences from"observing life"toward"computing life",laying a revolutionary technological foundation for future precision medicine,drug design,public health,and biomanufacturing.These directions synergistically advance the precision design of molecular structures,dynamic regulation of cellular behavior,and systematic reconstruction of tissue functions,ultimately enabling dynamic replication,precise intervention,and continuous optimization of physiological functions at the organ or even organismal level.Finally,the article emphasizes that,despite rapid technological convergence,challenges such as underdeveloped regulatory framework,ambiguous ethical boundaries,and poorly defined safety risks remain.Accordingly,it proposes the construction of a dynamic and adaptive regulatory system for convergent technologies,the establishment of new interdisciplinary"translation"and collaboration mechanisms,and the integration of ethics and safety considerations into the very source of technological innovation.Looking ahead,with continued breakthroughs in AI,digital twins,advanced materials,and automated platforms,the convergence of biological and non-biological technologies is expected to advance higher levels of intelligent collaboration,driving a new wave of scientific and industrial revolution.
Driven by synergistic breakthroughs in chemical modifications,delivery technologies,and mechanistic understanding,RNA therapeutics—spearheaded by oligonucleotide therapeutics and mRNA technologies—has rapidly evolved from fundamental discoveries into a phase of extensive clinical translation.It now stands as a pivotal frontier in global biomedical innovation.This review systematically analyzes the progress in 2025 across foundational platforms,disease-targeting research,clinical development,and industrialization,while assessing future trends and offering strategic recommendations.At the technology level,sequence and structural design have become more refined,evolving from an empirical and rule-dependent approach to a new stage of AI-driven rational design.Advances in chemical modification technologies targeting the backbone,ribose,and nucleobases continue to enhance RNA stability and effectively reduce immunogenicity,while achieving precise and intelligent modulation has emerged as a focal point in the field.Although current delivery systems remain dominated by liver-targeting approaches using lipid nanoparticles(LNPs)and GalNAc conjugates,next-generation technologies are diversifying toward overcoming hepatic clearance and achieving tissue-specific targeting.Meanwhile,the application of artificial intelligence tools is significantly accelerating the processes of sequence design and delivery vehicle screening.In the field of basic and clinical research,the application of oligonucleotide therapeutics is rapidly expanding from genetic disorders to chronic diseases such as cardiovascular and neurodegenerative conditions.mRNA technology has also advanced beyond infectious diseases into oncology and metabolic disorders,with particularly deepening progress in cancer immunotherapy,where personalized neoantigen vaccines have achieved milestone developments.Capitalizing on its inherent stability,circular RNA(circRNA)has emerged as a highly promising platform for novel vaccines and therapeutic drugs.As these technologies mature,the industry is transitioning from the proof-of-concept phase to commercial-scale production.Supported by innovative platform technologies and global collaborations,China is rising as a pivotal force in global biomedical innovation.Looking ahead,critical challenges remain in achieving extrahepatic tissue-specific delivery,realizing precise and tunable regulation,and understanding long-term biological effects.Overcoming these hurdles is essential for developing the next generation of accurate,safe,and broadly applicable RNA therapies.
Metabolic disorders comprise a group of diseases characterized by disturbances in energy metabolism,including major public health challenges such as obesity,type 2 diabetes,and metabolic dysfunction-associated steatotic liver disease(MASLD).Spermidine,a ubiquitous,naturally occurring polyamine,is essential for maintaining physiological health and sustaining metabolic balance.In the context of ageing research,spermidine has been shown to exert diverse and significant biological effects.Recent findings have increased interest in its capacity to improve metabolic health.Epidemiological evidence indicates a strong inverse association between dietary spermidine intake and the incidence of diabetes and cardiovascular diseases.Notably,spermidine concentrations decline progressively with age,particularly in elderly populations with a high prevalence of metabolic disorders.However,the mechanisms by which spermidine influences metabolic disease,as well as its therapeutic potential,remain incompletely understood.This review aims to systematically examine the biological roles of spermidine in metabolic disorders and elucidate the molecular mechanisms involved,providing a foundation for novel preventive and therapeutic strategies.This review first outlines the biosynthetic and transport pathways of spermidine,establishing a framework for understanding the origins and regulation of its biological functions.In vivo,the mammalian spermidine pool is derived from three primary sources:dietary intake,de novo biosynthesis,and gut microbiota-derived production.Its distribution and regulation rely on coordinated enzymatic cascades and membrane transport systems.The review then summarizes and discusses the key molecular mechanisms underlying spermidine's putative protective actions,which include:stimulation of autophagy to remove misfolded proteins and damaged organelles;serving as a substrate for eIF5A hypusination,thereby modulating translation and cell proliferation;suppression of pro-inflammatory signaling pathways such as NF-κB to reduce tissue inflammation;inhibition of apoptotic process and reinforcement of antioxidant defense to promote cell survival;and optimization of mitochondrial function to enhance metabolic efficiency.These interconnected processes establish spermidine as a critical regulator of metabolic stability.Through these mechanisms,spermidine demonstrates broad prophylactic and therapeutic potential across a spectrum of metabolic disorders.In obesity,it promotes weight management by activating adipose tissue lipolysis,enhancing thermogenesis in brown adipose tissue,and improving gut microbiota composition and intestinal barrier function.In diabetes,it contributes to glycaemic control by improving peripheral insulin sensitivity,preserving pancreatic β-cell function,and alleviating vascular endothelial inflammatory injury.In MASLD,it mitigates hepatic steatosis and inflammation by enhancing fatty acid oxidation,reducing lipid accumulation,and directly improving mitochondrial function,thereby retarding the progression of fibrosis.In cardiovascular diseases,it confers protection against hypertension and atherosclerosis through multiple pathways,including inducing cardiomyocyte autophagy to improve cardiac function and elasticity,modulating blood lipid profiles,inhibiting platelet aggregation,and stabilizing atherosclerotic plaques.In conclusion,spermidine supports metabolic homeostasis through a multifaceted,synergistic network of molecular actions,offering promising avenues for intervention in multiple metabolic disorders,as evidenced by preclinical research.However,its effects are influenced by cellular context and environmental factors,and translation to clinical practice is challenged by limited human trial data,uncertain dose-response parameters,and potential context-dependent adverse effects.Future advancements will require robust clinical studies informed by a comprehensive understanding of its complex molecular interactions and systems-level biology,with the ultimate goal of defining optimal therapeutic parameters and advancing spermidine as a viable strategy for metabolic disease prevention and treatment.
