The ’Endgame’ and ’Rebirth’ of Diagnostic Antibodies: A Scientific Leap from Discovery to Design

The ’Endgame’ and ’Rebirth’ of Diagnostic Antibodies: A Scientific Leap from Discovery to Design | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL View This is a preprint and has not been peer reviewed. Data may be preliminary. 15 December 2025 V1 Latest version Share on The ’Endgame’ and ’Rebirth’ of Diagnostic Antibodies: A Scientific Leap from Discovery to Design Authors : Qitao Song , Lulai Xu , Fei Xie , Qiang Li , Tiantian Yang , Xiaoxia Xu , Yanyin Lin , … Show All … , Lihua Zhang , Jiahui Zhou , Junyou Zhou , Linping Li , Chengdong Ji , Kejun Liu , Zhanhui Wang , Weijing Yi , and Jing Wang 0000-0001-7589-1862 [email protected] Show Fewer Authors Info & Affiliations https://doi.org/10.22541/au.176583441.18327723/v1 347 views 172 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract With the maturation and standardization of antibody discovery technologies, the core competitive advantage of diagnostic antibodies is shifting from discovery to design. Traditionally, antibody development has emphasized molecular recognition performance, with value defined by static parameters such as affinity, specificity, and stability. However, the emergence of mass spectrometry–based reverse development, deep single B-cell screening, and AI-driven antibody design has transformed antibody identification into a standardized capability. As a result, antibodies are no longer scarce resources, and their differentiated value is reflected in their behavioral attributes rather than mere recognition strength. Here, we propose a new paradigm in which diagnostic antibodies evolve from passive molecular recognition elements into structurally intelligent units. In this framework, antibody capabilities shift from being “stronger” to being “smarter, programmable, and capable of expressing system-level behaviors.” We define four key ability dimensions underlying this transformation: affinity stratification, conformational logic, functional coupling, and system behavior. Through practical examples, we demonstrate that the future value of diagnostic antibodies lies in their capacity to actively shape diagnostic systems, rather than simply participate as interchangeable reagents. As antibodies progressively acquire intrinsic decision-making logic and structural control properties, they will emerge as the primary intelligent units of in vitro diagnostic systems. Consequently, the ultimate goal of diagnostic antibody development is no longer the infinite optimization of recognition efficiency, the rational design of structural logic and functional programmability. This paradigm shift will fundamentally reshape diagnostic technologies and drive clinical testing from quantitative signal interpretation toward an emerging discipline of in vitro immune cognition. The ’Endgame’ and ’Rebirth’ of Diagnostic Antibodies: A Scientific Leap from Discovery to Design Qitao Song a, b, f, 1 , Lulai Xu a, b, c, 1 , Fei Xie g, 1 , Qiang Li b, 1 , Tiantian Yang f , Xiaoxia Xu a , Yanyin Lin a , Lihua Zhang a , Jiahui Zhou a, b , Junyou Zhou e , Linping Li e , Chengdong Ji b , Kejun Liu h, * , Zhanhui Wang d, * , Weijing Yi a, b, e* , Jing Wang a, b, e* a. Chongqing Essence Biological Engineering, Chongqing, China, 400082 b. Zybio Inc, Chongqing, 400082, China c. Animal Disease Prevention and Green Development Key Laboratory of Sichuan Province, College of Life Science, Sichuan University, Chengdu 610064, China; d. Beijing Advanced Innovation Center for Food Nutrition and Human Health, Beijing Laboratory of Food Quality and Safety, China Agricultural University, Beijing 100193, China e. School of Pharmacy and Bioengineering, Chongqing University of Technology, Chongqing, 400054 f. The Center for Clinical Molecular Medical Detection, The First Affiliated Hospital of Chongqing Medical University, Chongqing 400016, P. R. China g. Department of Outpatient, Taihe Hospital, Hubei University of Medicine, No 32 Renmin Road, Shiyan 442000, P. R. China h. National Institutes for Food and Drug Control, No. 31 Huatuo Road, Daxing District, Beijing 102629, China 1. Qitao Song, Lulai Xu, Fei Xie and Qiang Li contributed equally to this work. Corresponding Author Jing Wang (Email: [email protected] ), Weijing Yi (Email: [email protected] ), Zhanhui Wang (Email: [email protected] ), Kejun Liu ( [email protected] ) Abstract: With the maturation and standardization of antibody discovery technologies, the core competitive advantage of diagnostic antibodies is shifting from discovery to design. Traditionally, antibody development has emphasized molecular recognition performance, with value defined by static parameters such as affinity, specificity, and stability. However, the emergence of mass spectrometry–based reverse development, deep single B-cell screening, and AI-driven antibody design has transformed antibody identification into a standardized capability. As a result, antibodies are no longer scarce resources, and their differentiated value is reflected in their behavioral attributes rather than mere recognition strength. Here, we propose a new paradigm in which diagnostic antibodies evolve from passive molecular recognition elements into structurally intelligent units. In this framework, antibody capabilities shift from being “stronger” to being “smarter, programmable, and capable of expressing system-level behaviors.” We define four key ability dimensions underlying this transformation: affinity stratification, conformational logic, functional coupling, and system behavior. Through practical examples, we demonstrate that the future value of diagnostic antibodies lies in their capacity to actively shape diagnostic systems, rather than simply participate as interchangeable reagents. As antibodies progressively acquire intrinsic decision-making logic and structural control properties, they will emerge as the primary intelligent units of in vitro diagnostic systems. Consequently, the ultimate goal of diagnostic antibody development is no longer the infinite optimization of recognition efficiency, the rational design of structural logic and functional programmability. This paradigm shift will fundamentally reshape diagnostic technologies and drive clinical testing from quantitative signal interpretation toward an emerging discipline of in vitro immune cognition. Keywords: Diagnostic Antibody; Antibody Discovery; Antibody Design; In Vitro Diagnostics (IVD); Antibody Intelligence Introduction: The history of antibody technology has always been centered around the expansion of its ”capacity boundaries.” In 1890, Behring and Kitasato discovered ”antitoxins” in the serum of rabbits infected with tetanus, marking the first evidence that immunity could be transmitted through humoral factors, independent of cells [1]. In 1923, Heidelberger and Avery determined through chemical methods that antibodies are proteins, laying the conceptual foundation for ”molecular recognition” [2]. In 1975, Köhler and Milstein reduced ”serum heterogeneity” to ”monoclonal antibodies” through hybridoma technology, initiating the industrial era of diagnostic antibodies [3]. In 1990, McCafferty demonstrated the use of phage display technology to present antibody variable regions on the surface of phages, signaling the dawn of the paradigm where ”sequences are the raw material” [4]. Over the past four decades, this high-throughput discovery paradigm has made diagnostic antibodies the core and irreplaceable ”chip-level” raw material in the IVD industry. However, the overwhelming success of this paradigm has also sowed the seeds for its eventual obsolescence. As single B-cell screening now allows for the output of hundreds of candidate antibodies in one go [5], as protein reverse sequencing can replicate mature molecules within 72 hours [6], and as generative AI can design completely novel sequences from scratch in the cloud [7], ”discovery” itself has become an industrial process that can be scaled up and replicated. The once-prized metrics of ”high affinity” and ”high specificity” are now easily achieved by dozens of candidates targeting the same epitope, leading to rapidly diminishing value differences and a cycle of ”homogenization—price wars—further homogenization.” The ”discovery era” of diagnostic antibodies is approaching its end. Yet, this end does not signify ”death,” but rather a redefinition of ”capacity boundaries.” In the past three years, 42% of R&D investments by the global top 20 IVD companies have shifted from ”sequence screening” to ”function enhancement.” Some companies have transformed the Fc region into a pH-dependent switch [8], enabling homogeneous chemiluminescence assays; others have fused antibodies with DNAzymes [9, 10], achieving ”one-step colorimetric” detection of various analytes; and some domestic startups have embedded antibody heavy chains into quantum dot lattices for room-temperature single-molecule imaging of cardiac troponin I. These examples point to a hidden yet profound paradigm shift—antibodies are evolving from ”recognition molecules” to ”intelligent units,” where their core value is no longer simply ”binding,” but a system-level ability encompassing ”sensing—computing—storing—communicating.” At the same time, the evaluation criteria for clinical immunoassays are undergoing a significant shift. In complex samples (such as whole blood, saliva, and exhaled condensates), antibodies must not only ”recognize,” but also ”recognize correctly,” i.e., exhibiting ”tolerance-correction” capabilities to account for matrix interference, post-translational modifications, and conformational variations. In ultra-low concentration ranges (fg mL⁻¹ or even aM), antibodies must not only ”detect” but also ”express true clinical meaning”—signal outputs must correlate quantitatively with disease stages, prognosis, or therapeutic windows. In new methodological systems such as homogeneous chemiluminescence, real-time immunoassays, and wearable sensors, antibodies are no longer just ”compatible,” but must ”drive correct system behaviors”—acting as ”logic gates” that trigger chain reactions, self-assembly, signal amplification, or termination. These demands signify a crucial leap: antibodies are no longer ”parts of a detection system,” but are now positioned as ”the underlying logic of diagnostic behaviors,” with their role evolving from ”raw materials” to ”intelligent structures,” their attributes transitioning from ”binding” to ”behavior,” and their value shifting from ”discovery” to ”design.” In other words, the true watershed for diagnostic antibodies is no longer a technical breakthrough but a paradigm reconstruction—from ”discovery-based antibodies” to ”design-based antibodies.” In this new paradigm, the core value of antibodies is no longer about ”binding with what affinity,” but about ”whether they can carry structural logic and programmable behavior”: Can they reversibly switch in the pH range of 6.0 to 8.0? Can they release DNA tags upon antigen binding? Can they generate spin decoherence at the interface of quantum sensors? Can they amplify signals within cells and trigger downstream gene expression? These abilities cannot be achieved through ”discovery” alone; they require ”design”—i.e., the top-down programming of antibody sequences, conformations, modifications, conjugations, microenvironments, and even system interfaces. Therefore, this paper presents the following judgment: The marginal returns of the traditional ”discovery” route have approached zero, and the ”discovery era” of diagnostic antibodies is drawing to a close. A new paradigm centered around ”programmable behavior” is emerging, marking the ”birth” of a new phase. The value chain will shift from ”selling sequences” to ”selling functions,” from ”selling reagents” to ”selling decisions,” and from ”antibody companies” to ”antibody system operators.” It is hoped that this ”birth” after the ”end” will bring sustainable incremental value to clinical diagnostics, rather than an unbearable technological backlash. 1. Historical Transition and Technological Breakthroughs in Antibody Discovery: From Forward to Reverse Approaches 1.1. Historical Stages of Antibody Forward Discovery and its Technological Challenges and Breakthroughs Serum Era (1890s–1960s): Discovery of Antibodies and the Use of Polyclonal Antibodies. In the late 19th century, Emil Adolf von Behring and Shibasaburo Kitasato first discovered antitoxins through experiments involving tetanus-infected rabbit serum, laying the foundation for antibody research [1]. Subsequently, antibodies were identified as γ-globulins in serum [11], and the theory of antigen-antibody specific binding was proposed [12, 13]. However, during this stage, antibodies were primarily derived from immune animal serum, which resulted in polyclonal antibodies that suffered from significant batch-to-batch variation, poor specificity, and frequent cross-reactions, rendering them inadequate for high-precision diagnostics [14]. Despite these challenges, the serum era established antibodies as the fundamental molecules of immune recognition and provided the theoretical basis for the subsequent development of monoclonal antibodies ( Figure 1a ). Hybridoma Era (1975–1990s): Industrial Production of Monoclonal Antibodies. In 1975, Georges Köhler and César Milstein developed the hybridoma technology, which allowed for the immortal in vitro production of monoclonal antibodies by fusing B cells with myeloma cells [15]. This breakthrough resolved the issue of poor specificity inherent in polyclonal antibodies and greatly expanded the application of antibodies in diagnostics and therapeutics. However, hybridoma technology also faced limitations such as low cell fusion efficiency, lengthy screening processes, chromosomal instability in hybridoma cells, and high immunogenicity of murine-derived antibodies, which hindered their further clinical use [16]. Nevertheless, hybridoma technology facilitated the standardization and large-scale production of antibodies, marking the industrialization of antibody research [17] ( Figure 1b ). Phage Display Era (1990s–2010s): Rise of In Vitro Screening Technologies. With the development of molecular biology, phage display technology emerged, enabling the display of antibody variable regions (e.g., scFv or Fab) on the surface of bacteriophages for in vitro high-throughput screening [4, 18, 19]. This technology bypassed the cumbersome processes of animal immunization and cell fusion, significantly improving the efficiency of antibody discovery and driving the development of fully human antibody therapeutics [18]. However, phage display technology also faced challenges such as limited display fragment sizes, difficulty in retaining natural heavy-light chain pairings, the complexity of multi-round selection processes, and restricted antibody diversity [20]. Moreover, the technology required high-quality immunogens, and the development of antibodies against difficult-to-express targets like membrane proteins remained challenging [20]. Nevertheless, phage display facilitated the transition from ”cell-based screening” to ”gene-based screening” in antibody research, laying the foundation for future high-throughput screening technologies ( Figure 1c ). Single B-Cell Era (2010s–2020s): Breakthroughs in High-Throughput Screening Technologies. In recent years, rapid advancements in microfluidics, nanopores, and single-cell sequencing technologies have made single B-cell screening a key method for antibody discovery [5, 21]. This technique allows for the direct isolation of individual B cells from immune animals or humans, obtaining their antibody gene sequences for high-throughput, high-fidelity screening. Compared to traditional methods, single B-cell screening has significantly improved the efficiency and diversity of antibody discovery, providing a technological pathway for the development of high-performance antibodies, such as rabbit monoclonal and fully human antibodies. However, challenges remain, including the large number of candidate sequences, high costs for subsequent validation, dependence on high-expression cell lines (which are still imported in many countries), and the issue of antibody sequence homogeneity. Furthermore, the technique remains highly dependent on the quality of the immunogen and immunization strategy. Nevertheless, the single B-cell era marks a shift from ”low-throughput screening” to ”high-throughput discovery,” offering new possibilities for the precise development of antibodies ( Figure 1d ). Intelligent Design Era (2020s–Present): AI-Driven Paradigm Shift in Antibody Design. With the rapid development of artificial intelligence (AI) and computational biology, antibody research is progressively entering the era of intelligent design [6, 22]. This stage is characterized by AI-generated antibodies, computational antibody design, and structural-functional customization, overcoming the limitations of traditional ”discovery” models and enabling a paradigm shift from ”passive recognition” to ”active design” in antibody development [23]. Through AI algorithms, researchers can design antibodies with specific affinities, conformational stability, and functional properties in vitro, significantly shortening development cycles and reducing costs. However, challenges persist, such as insufficient experimental validation of AI-generated antibodies, the complex relationship between antibody structure and function, incomplete design rules, and the industry’s failure to shift from focusing on ”sequence value” to ”functional value” [24]. Additionally, regulatory and ethical frameworks have not yet fully adapted to the rapid development of ”designed antibodies.” Despite these hurdles, the intelligent design era represents a shift from ”experience-driven” to ”data-driven” antibody development. Antibodies are evolving from ”recognition molecules” to ”intelligent structural units,” providing new technological support for the intelligent and personalized development of in vitro diagnostic systems ( Figure 1e ). 1.2. The ”Reverse Discovery” Phase Opened by Protein Sequencing and its Technological Challenges and Breakthroughs Following the establishment of ”gene-to-function” forward discovery routes through hybridoma, phage display, and single B-cell technologies, mass spectrometry-based direct protein sequencing (de novo protein sequencing) was long considered a supplementary method and remained on the periphery [7]. This approach operates contrary to the ”DNA→RNA→protein” paradigm by directly proteolyzing and mass spectrometry-analyzing existing functional antibody molecules in serum, ascites, or affinity-purified products, bypassing the need to pre-acquire spleen cells from immune animals or construct antibody libraries. This ”terminal perspective” was systematically overlooked during the discovery era, primarily due to early limitations in mass spectrometry throughput, complex fragmentation patterns, and the lack of robust assembly algorithms, with the industry often equating ”sequence ownership” with ”successful gene cloning.” With the maturation of high-resolution mass spectrometry, stepped-energy fragmentation (HCD/ETD), and deep learning-based analysis algorithms (e.g., pNovo, DeepNovo), protein sequencing has enabled the reconstruction of multiple full-length monoclonal sequences from as little as 0.1 mg of mixed polyclonal antibodies within 48 hours, facilitating ”polyclonal-to-monoclonal conversion” [20]. Its key advantages include: (1) bypassing the need for immunogen preparation and animal immunization, directly capturing in vivo-validated antibody responses, which is particularly advantageous for membrane proteins and post-translationally modified epitopes that are difficult to express; (2) enabling the analysis of natural light-heavy chain pairings, preserving in vivo affinity maturation and glycosylation states, thus reducing the risk of mismatches during in vitro replication; (3) when coupled with computational platforms, allowing the simultaneous analysis of dozens of functional antibodies within a polyclonal mixture, rapidly constructing ”mini-panels” and significantly shortening diagnostic antibody combination development timelines. However, inherent limitations of terminal sequencing must not be overlooked: polyclonal antibody abundance in serum is dynamic and wide-ranging, with low-copy antibodies being easily missed; chemical modifications (e.g., oxidation, deamidation) introduce noise in spectra, leading to reduced sequence coverage; distinguishing functional from non-functional background antibodies remains a challenge, requiring downstream recombinant expression and epitope validation. Moreover, protein sequencing bypasses genetic intellectual property markers, which could lead to sequence traceability and patent disputes. Therefore, protein sequencing should not be seen as a ”shortcut” to replace forward discovery but rather as a ”reverse engineering” tool to fill discovery gaps. In scenarios involving urgent outbreaks, limited samples, or highly mature natural immune responses, combining protein sequencing with single B-cell and AI-driven approaches forms a ”bi-directional complementary” strategy that maximizes the value of diagnostic antibody development ( Figure 2 ). 2. The Emergence and Future Development Trends of the “Antibody Design Era” The transformation towards antibody design is not merely an extension of research paradigms, but also represents a systematic reconstruction of molecular diagnostics. From “passive discovery” to “active definition,” the research focus on antibodies is shifting from simple recognition performance to structural logic, behavioral plasticity, and system intelligence. Future antibody research will no longer be confined to sequence screening or affinity optimization but will enter a new phase centered on multi-level structural programming and systemic behavior regulation. In this phase, antibodies will become “structural intelligence units” of in vitro diagnostic systems, possessing adaptability, information feedback, and system regulatory functions. This will lead to a fundamental leap in diagnostic systems, transitioning from “accurate measurement” to “deep understanding” [25, 26]. 2.1 Antibody Programmable Conformational Design Based on Structural Dynamics The primary direction of future antibody design is the programming of conformational dynamics. Traditional antibodies have been viewed as static structures; however, recent studies indicate that the spatial dynamics between the variable regions (VH, VL) play a central role in determining binding precision and system stability. By introducing adjustable peptide bridges, reversible conformational locks, or pH-responsive amino acid residues, antibodies can adopt specific conformations at different stages of the reaction. For instance, the Stanford University team developed a “pH-switchable antibody” that alters the binding site orientation through histidine modification under acidic conditions, enabling automatic switching between antigen binding and dissociation [26]. This mechanism can avoid high background signals and nonspecific adsorption in clinical diagnostics, optimizing dynamic signals. Similar strategies have been applied by domestic research teams in homogeneous immunoassay systems, developing antibodies that spontaneously adjust binding kinetics in liquid-phase systems [27], showcasing potential applications in complex diagnostic scenarios. These studies suggest that future antibody design will shift from “fixed affinity” to “tunable behavior,” using structural dynamics models to predict and control antibody response behavior [25] ( Figure 3a ). 2.2 Antibody-System Integrated Design with a Focus on Multifunctional Fusion A second development trend is the fusion of functions and the co-design of antibody systems. In traditional immunoassays, signal amplification typically relies on exogenous labeled enzymes or fluorophores, but future design concepts emphasize embedding the signal functionality directly into the antibody molecule. Recent studies have successfully constructed antibody-enzyme fusion proteins (antibody-enzyme fusions), such as the natural fusion of antibodies with alkaline phosphatase (ALP), horseradish peroxidase (HRP), or fluorescent proteins. These molecules can complete both recognition and signal amplification in a single expression system. For example, a joint team from the University of Tokyo and Roche developed an “ALP-fused antibody” that achieved unlabelled amplification in a chemiluminescence system, improving detection sensitivity by approximately 10 times. Some companies are also exploring fusion diagnostic antibodies, such as incorporating directional fusion tags (Dock-tag) and small molecule substrate co-modules in chemiluminescence systems, realizing system-integrated signal responses [28-30]. The significance of this strategy is that antibodies are no longer just components of a detection system but become carriers of the system’s logic, integrating “recognition—signal—regulation” functions, laying the molecular foundation for intelligent diagnostic systems ( Figure 3b ). 2.3 AI-Driven Antibody Generation and Behavior Prediction Models The third critical direction is the deep integration of artificial intelligence (AI) into antibody design. With the development of protein language models (Protein Language Models) and deep structural generation networks, antibody sequence design has transitioned from an experience-based screening process to an algorithm-driven prediction process. Currently, models such as AlphaBind, ESM-2, and Ig-VAE can directly generate feasible antibody sequences based on antigen structures, predicting their conformational stability and affinity trends through energy evaluation and binding interface analysis [31-34]. The introduction of AI technology not only significantly shortens the R&D cycle but, more importantly, drives “behavioral-level” design: models can learn the performance patterns of different antibodies in a system, thereby generating molecules with target behavioral features. Some studies have used Generative Adversarial Networks (GANs) to generate “specific curve response antibodies,” allowing antibodies to exhibit linear responses within a defined concentration range [35]. This AI-based antibody development model, which starts from “designing behavior,” marks the entry of antibody development into a “data-driven, experimental validation” closed-loop system and signals the technical realization of the “intelligent antibody” concept ( Figure 3c ). 2.4 Modular Antibody Framework and Diagnostic System Restructuring As the applications of antibodies expand, modular design is becoming the key path for systematic antibody construction. Future antibodies may be divided into standardized modules, such as recognition units, amplification units, signal regulation units, and stability control units, with modular combinations enabling rapid functional reconfiguration. For example, the Massachusetts Institute of Technology (MIT) proposed an “antibody logic module” system, where different domains are modularly spliced to achieve multi-condition triggered recognition, enabling selective antigen recognition under specific metabolic states. In diagnostic fields, this design concept can be applied to develop “multi-condition response antibodies,” improving discrimination accuracy in multi-analyte co-detection [36-38]. Some domestic companies have attempted to develop universal antibody modules that can switch between different detection platforms based on modular reconfiguration strategies, such as sharing recognition domains between immunoturbidimetry and chemiluminescence systems. This “platform-adaptive design” indicates that future antibodies will become universal programming interfaces for diagnostic systems ( Figure 3d ). 2.5 Future Outlook: From Structural Intelligence Units to Diagnostic Cognitive Units Based on the trends outlined above, it is anticipated that the future development of antibodies will follow the overall direction of “structural intelligence—system integration—algorithm-driven.” Future antibodies will no longer simply be materials for signal recognition, but will become diagnostic intelligent units with “logic,” “programmability,” and “system interactivity.” Their design principles will shift from single performance indicators to cross-dimensional capability optimization, including tunability of recognition, self-correction of signals, and system collaboration [7, 39]. With the integration of AI models, molecular simulations, and high-throughput experimentation, antibodies will exhibit “trainability, controllability, and interpretability,” becoming the core driving force for diagnostic intelligence. From a scientific philosophical perspective, this evolution signifies that antibodies have transitioned from “the final product of natural selection” to “artificially defined functional entities,” ushering in a new era where diagnostics will move from “passive perception” to “active cognition” ( Figure 3e ). 3. From ”Molecular Discovery” to ”Cognitive Construction”: A Scientific Leap The development of diagnostic antibodies is a scientific history of continuous self-updating technological paradigms. From serum antitoxins to monoclonal antibodies, from display libraries to single B-cell screening, and now to AI-driven design, each innovation has extended the boundaries of human understanding of the logic of biological recognition. Looking back at over a century of evolution, it becomes clear that the ”endgame” of antibody development is not technological stagnation but a transformation of the ”problem itself”—when the efficiency of discovery approaches its limit, science shifts toward higher-level questions: Can antibodies possess the ability to ”understand” and ”control” system behaviors? In other words, the focus of antibody research is shifting from ”how to find binding” to ”how to design meaning.” This shift is not only a methodological update but also a migration of scientific cognition. Traditional antibody development relied on biological screening and empirical accumulation, with its core logic being ”extracting answers from nature.” However, the emergence of designer antibodies allowed researchers for the first time to ”pose questions and embed answers”—by collaboratively modeling sequences, conformations, and system environments, antibodies became the smallest unit to execute diagnostic logic. This means that antibodies are no longer mere imitations of nature but rather the initial acquisition of ”design rights” over the immune system at the molecular level. Consequently, antibodies have evolved from being products of biological systems to platforms for realizing biological logic, becoming fundamental components of ”artificial immune intelligence.” Future diagnostic science may therefore witness a paradigm shift in cognition. With antibodies possessing programmability and logical expression capabilities, diagnostic systems will no longer rely solely on individual reaction signals but will shift toward an information perception, judgment, and feedback process collectively carried out by antibody clusters. The underlying logic of clinical testing will transform from an ”input-output” model to a ”perception-decision-adaptation” model, enabling diagnostic systems to exhibit feedback characteristics similar to biological cognition. This trend will drive in vitro diagnostics from a ”signal measurement science” to a ”cognitive engineering science,” centered on molecular behavior and logic, to build explainable and evolvable intelligent diagnostic systems ( Figure 4 ). From an industrial perspective, this transformation will also profoundly reshape the value chain. The future competitive focus will no longer be on the quantity of antibody sequences or production scale, but on whether antibodies can play the role of ”cognitive nodes” at the system level. Antibody companies will shift from being ”material providers” to ”intelligent system designers”; diagnostic products will evolve from ”reagent kits” to ”structured knowledge carriers”; and testing behaviors will transition from ”passive measurement” to ”active judgment.” The essence of this transformation is a revolution in the way knowledge is produced: antibodies will no longer be a natural resource but rather an intelligent medium that can be arranged, restructured, and optimized by humans. Therefore, the ”endgame” of diagnostic antibodies is not the end of discovery but the beginning of understanding; its ”rebirth” is not the emergence of new sequences but the generation of new logic. When antibodies rise from passive molecular recognition to cognitive units with ”structural logic,” what they carry is no longer just diagnostic function but humanity’s first conscious construction of an intelligent system at the molecular level. In the future, in vitro diagnostics will no longer just be a tool for detecting disease but will become an experimental platform for humans to understand the laws of life and reconstruct immune cognition. This is the true rebirth of antibody science, transitioning from ”molecular discovery” to ”cognitive construction.” Notes The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This work is financially supported by New Chongqing Youth Innovation Talents Program (CSTB2024NSCQ-QCXMX0006, CSTB2025YITP-QCRCX0027, CSTB2025YITP-QCRCX0028) and Major Special Project for Technology Innovation and Application Development of Chongqing (CSTB2023TIAD-STX0011), Chongqing Natural Science Foundation General Project (2025NSCQ-GPX0680); Project supported by the National Natural Science Foundation of China (No.2250070464). References: [1] E. Behring, S. Kitasato, Dtsch Med Wochenschrift 1890, 16, 1113. [2] M. Heidelberger and O. T. Avery, J Exp Med, 1923, 38, 0073-0079. [3] G. Kohler and C. Milstein, Nature, 1975, 256, 495-497. [4] J. Mccafferty, A. D. Griffiths, G. Winter and D. J. Chiswell, Nature, 1990, 348, 552-554. [5] T. Tiller, E. Meffre, S. Yurasov, M. Tsuiji, M. C. Nussenzweig and H. Wardemann, J Immunol Methods, 2008, 334, 142-142. [6] N. H. Tran, X. L. L. Zhang, L. Xin, B. Z. Shan and M. Li, P Natl Acad Sci USA, 2017, 114, 8247-8252. [7] J. E. Shin, A. J. Riesselman, A. W. Kollasch, C. McMahon, E. Simon, C. Sander, A. Manglik, A. C. Kruse and D. S. Marks, Nat Commun, 2021, 12. [8] J. Reusch, J. T. Andersen, U. Rant and T. Schlothauer, Mabs-Austin, 2024, 16. [9] N. R. Jo, K. J. Lee and Y. B. Shin, Biosens Bioelectron, 2016, 81, 324-333. [10] Z. Zhen, L. J. Tang, H. X. Long and J. H. Jiang, Analytical Chemistry, 2012, 84, 3614-3620. [11] A. Tiselius and E. A. Kabat, J Exp Med, 1939, 69, 119-131. [12] C. K. Ingold, Nature 1940, 145, 644. [13] L. Pauling, Cc/Eng Tech Appl Sci, 1985, 16-16. [14] S. L. Morrison and V. T. Oi, Advances in Immunology, 1989, 44, 65-92. [15] G. Kohler and C. Milstein, Nature, 1975, 256, 495-497. [16] J. Z. Moraes, B. Hamaguchi, C. Braggion, E. R. Speciale, F. B. V. Cesar, G. F. D. S. Soares, J. H. Osaki, T. M. Pereira, R. B. Aguiar, Current research in immunology, 2021, 2, 32. [17] M. Tomita, M. Miwa, B. Le, S. Miyajima and T. Y. Tsong, Protein Eng, 1996, 9, 22-22. [18] N. Scheinfeld, Expert opinion on drug safety, 2005, 4, 637-641. [19] J. Mccafferty, A. D. Griffiths, G. Winter and D. J. Chiswell, Nature, 1990, 348, 552-554. [20] J. Wang, Q. T. Song, T. Yang, Y. L. Li, L. H. Zhang, J. Y. Li, F. F. Liu, Y. Y. Lin, X. X. Xu, Y. Heng, L. L. Xu, S. Zhang, J. H. Zhou, Y. B. Liu, L. Y. Kong, D. B. Tang, C. D. Ji, B. Tan, P. Liao, N. K. Pan, W. J. Yi and Z. H. Wang, View-China, 2024, 5. [21] M. Huang, Y. Y. Ma, X. Y. Gao, X. Y. Li, Q. Ding, C. X. Liu, X. P. Liu, H. Zhang and N. B. Yang, J Oncol, 2021, 2021. [22] R. F. Service, Science, 2025, 388, 689-690. [23] N. R. Bennett, J. L. Watson, R. J. Ragotte, A. J. Borst, D. L. See, C. Weidle, R. Biswas, Y. T. Yu, E. L. Shrock, R. Ault, P. J. Y. Leung, B. W. Huang, I. Goreshnik, J. Tam, K. D. Carr, B. Singer, C. Criswell, B. I. M. Wicky, D. Vafeados, M. G. Sanchez, H. M. Kim, S. V. Torres, S. D. Chan, S. M. Sun, T. T. Spear, Y. Sun, K. O’Reilly, J. M. Maris, N. G. Sgourakis, R. A. Melnyk, C. C. Liu and D. Baker, Nature, 2025, DOI: 10.1038/s41586-025-09721-5. [24] E. Callaway, Nature. 2024, 10.1038/d41586-024-00846-7 [25] M. Yoshida and M. Oda, Molecules, 2025, 30. [26] L. Biewenga, R. Vermathen, B. J. H. M. Rosier and M. Merkx, Acs Chem Biol, 2023, 19, 48-57. [27] W. X. Xie, J. Chen, J. H. Fu, Z. Tu, L. Y. Duan, W. Y. Yang, X. T. Hu, Q. H. He, W. J. Wang and D. Wang, Analytical Chemistry, 2025, DOI: 10.1021/acs.analchem.5c05346. [28] F. N. Hitawala and J. J. Gray, Mabs-Austin, 2025, 17. [29] Y. Huang, Z. Z. Li, Q. Y. Hong, L. Z. Zhou, Y. Ma, Y. S. Hu, J. B. Xin, T. T. Li, Z. B. Kong, Q. B. Zheng, Y. X. Chen, Q. J. Zhao, Y. Gu, J. Zhang, Y. B. Wang, H. Yu, S. W. Li and N. S. Xia, Comput Struct Biotec, 2022, 20, 710-720. [30] C. Geuijen, D. Maussang-Detaille, T. Gallenne, L. Hendriks, A. Kramer, J. Gerlach, M. Throsby and J. de Kruif, Cancer Res, 2021, 81. [31] M. F. Chungyoun and J. J. Gray, Curr Opin Biomed Eng, 2023, 28. [32] A. A. Agarwal, J. Harrang, D. Noble, K. L. Mcgowan, A. W. Lange, E. Engelhart, M. C. Lahman, J. Adamo, X. Yu, O. Serang, K. J. Minch, K. Y. Wellman, D. A. Younger, R. M. Lopez and R. O. Emerson, Mabs-Austin, 2025, 17. [33] Z. Demir, M. Raynaud, G. Divard, K. Louis, A. Truchot, M. Niemann, R. Ponsirenas, O. Aubert, A. Del Bello, A. Hertig, D. Anglicheau, B. Dale, N. Kamar, M. Mangiola, A. Zeevi, C. Lefaucheur and A. Loupy, Nat Commun, 2025, 16. [34] R. R. Eguchi, C. A. Choe and P. S. Huang, Plos Comput Biol, 2022, 18. [35] D. J. Wang, Y. S. Ge, B. Zhong and D. Liu, Comput Struct Biotec, 2021, 19, 1661-1671. [36] Y. Zhou, Y. Y. Wei, J. Zhang, X. Y. Shi, L. Ma and R. Yuan, Analytical Chemistry, 2024, 96, 11326-11333. [37] A. M. M. Rosa, A. F. Louro, S. A. M. Martins, J. Inácio, A. M. Azevedo and D. M. F. Prazeres, Analytical Chemistry, 2014, 86, 4340-4347. [38] M. T. W. Lee, A. Maruani, D. A. Richards, J. R. Baker, S. Caddick and V. Chudasama, Chem Sci, 2017, 8, 2056-2060. [39] S. J. Kellmann, S. Dübel and H. Thie, Mabs-Austin, 2017, 9, 404-418. Figure 1. The historical leap of antibody discovery technologies. From the serum era (a) to the hybridoma era (b), the display technology era (c), the single B-cell era (d), and the intelligent design era (e), each technological breakthrough has driven the evolution of antibody discovery and application. Figure 2. The ”Reverse Discovery” phase opened by mass spectrometry-based protein sequencing and its technological breakthroughs. Figure 3. Future trends and applications of antibody design. (a) Programmable conformational design based on structural dynamics; (b) Antibody-system integrated design with an emphasis on functional fusion; (c) AI-driven antibody generation and behavior prediction models; (d) Modular antibody frameworks and diagnostic system restructuring; (e) Future outlook: From structural intelligence units to diagnostic cognitive units. Figure 4. The scientific leap from ”Molecular Discovery” to ”Cognitive Construction.” Antibodies transition from traditional ”recognition molecules” to cognitive units with ”structural logic,” driving diagnostic science from a ”signal measurement science” to a ”cognitive engineering science” in the new era. Information & Authors Information Version history V1 Version 1 15 December 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Collection View Keywords antibody design antibody discovery antibody intelligence diagnostic antibody in vitro diagnostics (ivd) Authors Affiliations Qitao Song Zybio Inc, Chongqing, View all articles by this author Lulai Xu Zybio Inc View all articles by this author Fei Xie Hubei University of Medicine Taihe Hospital View all articles by this author Qiang Li Zybio Inc View all articles by this author Tiantian Yang The First Affiliated Hospital of Chongqing Medical University View all articles by this author Xiaoxia Xu Chongqing Essence Biological Engineering View all articles by this author Yanyin Lin Chongqing Essence Biological Engineering View all articles by this author Lihua Zhang Zybio Inc View all articles by this author Jiahui Zhou Zybio Inc View all articles by this author Junyou Zhou Zybio Inc View all articles by this author Linping Li Zybio Inc View all articles by this author Chengdong Ji Zybio Inc View all articles by this author Kejun Liu National Institutes for Food and Drug Control View all articles by this author Zhanhui Wang China Agricultural University View all articles by this author Weijing Yi Zybio Inc View all articles by this author Jing Wang 0000-0001-7589-1862 [email protected] Peking University View all articles by this author Metrics & Citations Metrics Article Usage 347 views 172 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Qitao Song, Lulai Xu, Fei Xie, et al. The ’Endgame’ and ’Rebirth’ of Diagnostic Antibodies: A Scientific Leap from Discovery to Design. Authorea . 15 December 2025. DOI: https://doi.org/10.22541/au.176583441.18327723/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.176583441.18327723/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ffe41b4dedb8e2e',t:'MTc3OTQ3ODQ0OQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();