ºÚÁÏÉçÇø

Molecules are incredibly small and often structurally similar, and identifying them one by one has long been slow, complex and difficult to reproduce. The AI Enabled Nanopore Sensing team has developed an AI-empowered electrochemical sensing method to ‘listen’ to single molecules as they pass through a tiny pore one by one, converting their molecular features into clear and unique fingerprints.

The team has built an integrated nanopore sensing instrument that makes single-molecule measurements rapid, accurate and accessible beyond specialist laboratories. This system enables the analysis of peptides and nucleic acids at the single-molecule level, with the potential to support applications in proteomics and genomics. By making molecular analysis more reliable and practical, this work helps bring advanced single-molecule sensing into everyday laboratory and clinical settings, with potential applications in early disease diagnosis and precision medicine.

This prize is a meaningful recognition of our team’s dedication over the past decade to advancing nanopore electrochemistry.

Yi-Tao Long

Q&A

What are your feelings on receiving this prize? 

Yi-Tao Long: We are truly delighted and honoured to receive this prestigious prize. It is a meaningful recognition of our team’s dedication over the past decade to advance nanopore electrochemistry. What has continually inspired us is the unique capability of nanopores. Their confined and precisely controlled nanoscale channels enable sequential analyte-pore interactions, generating measurable electrical signals from individual molecules, particles, and cells. By functioning as nanoscale electrodes, nanopores allow us to probe interfacial dynamics and reactivity at soft solid-liquid interfaces with unprecedented precision at the single-entity level. It has been incredibly rewarding to witness and contribute to the growth of this field from an emerging concept into a vibrant interdisciplinary research area.

Yi-Lun Ying:  After a decade of research on nanopore sensing with machine learning, we are honored and proud to see our progress as we have witnessed how AI is transforming single-molecule technologies, making them more intelligent and accessible. It is gratifying to see this trajectory being increasingly recognized, and we are pleased to contribute to this evolving field.

What was your role within the team? 

Yi-Lun Ying: My work is to explore the implementation of nanopore technology into application-oriented and real-world instruments. Specifically, I focus on improving the reliability and reproducibility of the technology in practical settings. By integrating AI-empowered analysis, this system is being developed to support clinical applications and to enable emerging nanopore fields, such as glycan sequencing and single-molecule reaction studies. 

Yongjing Wan: My role within the team is to lead the development of nanopore single-molecule signal processing. My academic background is in artificial intelligence and information science, and I have been working on nanopore-related research for nearly 10 years. It has been both exciting and intellectually rewarding to integrate advanced concepts from information science into the analysis of nanopore single-molecule signals, enabling more robust and insightful interpretation of complex experimental data.

Rong Wang: My expertise is in smart instrumentation, and I contribute to the development of experimental instruments within the team. I am also very pleased to work in a highly interdisciplinary environment, where PhD students with backgrounds in chemistry are actively engaged in instrument development. This close collaboration between engineering and chemistry has been both productive and inspiring, allowing us to jointly advance the design and implementation of nanopore-based technologies.

Meng-Yin Li: My role has been to uncover the sensing mechanisms for nanopore-based single-protein sequencing, and to translate these insights into improved sensitivity and resolution. In particular, I contributed to establishing a volume-interaction synergistic sensing mechanism, which explains how steric effects and specific molecular interactions jointly determine nanopore events. Based on this understanding, I have been involved in designing interaction-enhanced nanopores to improve sensitivity and selectivity. These efforts have enabled the identification of protein post-translational modifications at the single-molecule level.

Lin-Lin Zhang: My work is to develop the integrated nanopore instrument. We focus on designing an AI-assisted nanopore platform capable of sensing in real samples with high stability, sensitivity, selectivity, and precision, while also enabling rapid, portable, and cost-effective detection. By incorporating integrated measurement circuits and automated systems, we establish a robust data acquisition framework that supports high-throughput and robust single-molecule datasets for AI-driven identification.

What were the biggest challenges in this project, and how did you overcome them?

Wei Liu: The biggest challenge of this technology lies in overcoming a major bottleneck in isomer detection. Our strategy integrates precise nanopore engineering with a ‘synthesis-by-sensing’ approach at the single-molecule level to enhance the stereochemical discrimination of peptide stereoisomers.

Cheng-Bing Zhong: The nanopore ionic current signal is extremely sensitive to environmental noise, temperature fluctuations and membrane fluctuations. During my PhD studies, I advanced nanopore electrochemical measurement systems from single-channel configurations to multiplexed arrays, and developed high-bandwidth, low-noise current amplifiers.

We developed a fast digital acquisition and control system tailored for nanopore sensing, overcoming limitations of current amplifiers in real-time feedback and multi-channel synchronisation. The system enables pA-level, low-noise current recording, fast ADC conversion, and synchronised control of voltage and fluidics, achieving microsecond temporal resolution with low-latency feedback.

Ying-Huan Fu: A key challenge was ensuring that machine learning models remained both accurate and physically interpretable. Single-molecule signals are inherently noisy and often highly similar, which makes purely data-driven approaches prone to overfitting. To address this, we developed a frequency-fingerprint framework derived from the nanopore ion–analyte interaction network, improving both model robustness and interpretability. Our AI model enables reliable biomolecule identification.

What different strengths did different people bring to the team? 

Yi-Tao Long: Our team brings together complementary expertise across chemistry, electronic engineering and artificial intelligence. Chemists and electrochemists design nanopores with controllable molecular interactions, enabling selective signal amplification at the single-molecule level. Electronic engineers develop high-bandwidth, low-current detection systems and overcome the inherent instability of such measurements, ensuring reliable signal acquisition. Meanwhile, AI and data science experts create advanced algorithms to interpret complex signals and extract meaningful molecular fingerprints. By combining these strengths, we are able to integrate single-molecule sensing, instrumentation and data analysis into a unified platform, enabling both fundamental discoveries and real-world applications.

Why is this work so important and exciting? 

Yi-Tao Long: Single-molecule techniques were once confined to highly specialized laboratories with strict experimental requirements. Through our team’s interdisciplinary efforts, nanopore-based single-molecule sensing has become far more accessible, now being used in multiple laboratories and even introduced into undergraduate teaching as early as the second year. This technology is no longer distant from everyday research and application. By improving stability and sensitivity, it is opening up new possibilities in rapid multi-omics screening, proteomics analysis and sequencing, glycomics, and the sensing of information-carrying biomolecules. We believe it will become possible to accurately read and identify every single molecule in solution, enabling a deeper understanding of biological systems and transforming how we diagnose diseases, develop medicines, and interact with molecular information.

Meng-Yin Li: What excites me most is that the nanopore technique allows us to directly ‘watch’ and quantify molecular processes at the single-molecule level in real time. In many biological and chemical systems, important information is hidden because ensemble measurements average over heterogeneous and dynamic behaviour. Nanopore measurements give us a way to directly access these transient and rare molecular events. By improving the reliability and usability of this technology, we are opening up a more direct window into complex molecular systems, with potential impact on both basic science and applications.

Where do you see the biggest impact of this technology/research being? 

Hui Ma: The biggest impact of this technology ranges from early disease detection to personalised drug screening. Many disease biomarkers are low-abundance proteins, such as p53 for cancer and tau for Alzheimer’s disease. Existing assays, such as ELISA, lack single-molecule sensitivity and speed. Our AI–nanopore platform could detect a single pathogenic protein among millions of normal ones.

Fan Gao: I think the biggest impact will be in areas requiring precise molecular identification, such as trace impurity quantification and drug quality control. By enabling analysis at the single-molecule level, this approach offers a powerful advantage in resolving heterogeneous samples, where conventional bulk methods often average out subtle differences. As a result, both sensitivity and specificity can be substantially enhanced, enabling effective sensing of ultra-low-abundance and structurally similar molecules.

How will this work be used in real life applications? 

Yi-Lun Ying: With advances in instrumentation, algorithms and nanopore interface design, the next step is to focus on real-world, application-driven development. This includes areas such as peptide biomarker detection and impurity analysis in peptide-based drugs, where rapid and sensitive molecular identification is critically needed. Within the next 5 to 10 years, it has the potential to evolve into a general and versatile tool in the life sciences, supporting applications across diagnostics, drug development and molecular analysis.

Lin-Lin Zhang: Nanopore technology has been widely applied in single-molecule sensing for DNA, RNA and peptides, but its application is limited to analytes with previously characterised signals, as molecular identification relies on matching measured signals to existing purified samples. By integrating AI and automated instrumentation, our approach enables the generation and analysis of large-scale datasets from purified samples, allowing the system to learn molecular signatures and extend recognition to previously unknown molecules. In practice, this capability can facilitate applications such as novel biomarker discovery and rapid point-of-care testing, with broader real-world applicability.

How do you see this work developing over the next few years, and what is next for this technology/research? 

Cheng-Bing Zhong: We envision a fully automated nanopore platform that integrates the entire workflow, from AI-assisted sequence library design and AI-guided nanopore engineering to single-molecule experiments and downstream data analysis. Through this closed-loop system, it would be possible to build a large-scale, continuously expanding database of nanopore single-molecule signals.

Such a framework could substantially expand the capabilities of nanopore technology, enabling sequence-level analysis beyond proteins and nucleic acids to synthetic polymers, lipids and glycans. In the longer term, this may open the door to achieving multi-omics analysis at the single-molecule level.

How important would you say collaboration is for producing high-quality science? How has collaboration influenced your work? 

Yi-Tao Long: Collaboration is fundamental to our work. Nanopore sensing requires the integration of chemistry, electronic engineering and artificial intelligence, and each part depends on the others. From the very beginning, we design our single-molecule interface strategies across the full pipeline, from molecular concepts to instruments and algorithms. This close collaboration enables us to move efficiently from fundamental ideas to real-world applications.

Fan Gao: Collaboration is very important for us as students. I have benefited greatly from working with people from different backgrounds and a wide range of expertise. It not only accelerates problem-solving but also broadens how we think about specific scientific questions. Many solutions in this project would not have been possible within a single research area.

What does good research culture mean to you, and why does it matter? 

Zheng-Li Hu: To me, good research culture means open communication, interdisciplinary exchange and collaborative progress. It is a culture where people are encouraged to share ideas, learn from one another and work together across disciplinary boundaries. This kind of environment often leads to new research directions, fresh discoveries and more creative solutions to complex problems. It matters because good research culture not only helps projects move forward more efficiently, but also fosters innovation, supports long-term growth and creates the conditions for meaningful scientific breakthroughs.

Juan Tang: Good research culture comes down to trust. It means you can share a half-baked idea without worrying about being judged. You know your colleagues have your back. It’s okay to admit you don’t have all the answers, and messing up an experiment is just part of the process, not a failure.

Wei Liu: Good research culture means an environment built on scientific integrity, open communication and collaboration. During my research, frequent international academic exchanges and the freedom to discuss ideas with group members and supervisors have greatly accelerated my growth. This open and supportive atmosphere encourages critical thinking, improves research quality and helps researchers continuously develop both scientifically and professionally.

What do you wish more people understood about your field or the chemical sciences in general? 

Zheng-Li Hu: I wish more people recognised the transformative potential of nanopore electrochemistry. By advancing methods and instrumentation for controllable confined-space measurements, this field is expanding beyond single-molecule analysis into diagnostics, environmental monitoring and cutting-edge applications such as single-molecule reactions and synthesis. The integration of AI for interface design and data analysis, along with coupling to mass spectrometry and optical technologies, further highlights its interdisciplinary nature and innovation potential.

Shaochuang Liu: I wish more people understood that modern chemical science is far more than the study of molecules. Many of today's breakthroughs emerge from the integration of chemistry with biology, materials science, physics, engineering, and data science. Nanopore electrochemistry is a good example. The nanopore itself may originate from biochemistry or advanced materials, the measurements rely on analytical chemistry and electrochemistry, the experiments require sophisticated instrumentation, and the resulting data often depend on artificial intelligence. It is the convergence of these disciplines that gives us the remarkable ability to observe and understand individual molecules one at a time. For me, this is the true power of modern chemical science, not only understanding matter, but also creating entirely new ways to observe it.

Thinking back to earlier in your career, are there any words of wisdom that you wish someone had told you? 

Yi-Lun Ying: Looking back, some of the most valuable experiences in my career came from stepping outside my comfort zone. As an electrochemist, exploring areas beyond my original training has been challenging at times, but it is exactly this cross-disciplinary work that has led to new ideas and directions.

Hui Ma: â€˜Noise is not your enemy. It’s just information you haven’t learned to read yet.’ If there is one piece of advice I wish I had understood earlier, it is that fluctuations are not simply experimental imperfections. They are often the most information-rich component of the measurement, especially in single-molecule systems. This fundamentally changed how I interpret experimental data and think about what constitutes a meaningful signal.

Jie Jiang: One piece of wisdom I wish someone had told me earlier is that, once a problem is truly defined, the solution often begins to reveal itself. Over time, I have come to realise that in research, defining the question is not just the starting point, but often the most critical and creative part of the entire process. Many moments of confusion or stagnation do not come from a lack of ability to solve the problem, but from not yet seeing the essence of the problem clearly enough.

Juan Tang: I have learned that you cannot go a thousand miles without taking every single step. For a scientist, that means a strong foundation is essential. It is the only way your work can be both deep and lasting.