Prof. Bairong Shen
Professor and director general of the Institutes for Systems Genetics at West China Hospital

Sichuan University, China

Bairong Shen received his PhD degree in Physical Chemistry from Fudan University in 1997. He began his research in bioinformatics in June 1999 and underwent postdoctoral training at the University of Tampere, Finland. Following that, he was recruited as an assistant/associate professor of systems biology in early 2004. In June 2008, he returned to China and established the Center for Systems Biology at Soochow University, where he served as the director. In the summer of 2018, he was appointed as a professor and the director general of the Institutes for Systems Genetics at West China Hospital, Sichuan University, China. Throughout the past 25 years, he has published over 300 scientific papers and 10 books. He has also served as a peer reviewer for more than 30 international journals and is an affiliated faculty member at the Institute for Systems Biology in Seattle. Additionally, he is the founding chair of the International Conference on Translational Informatics (ICTI). His research interests encompass biomarker discovery, translational informatics, and smart healthcare.

 

Speech Title: "Vertical LLM Application, Horizontal Small-Model Expansion, and Human-AI Synergy in Smart Medicine"

 

Abstract: Large language models (LLMs) face inherent bottlenecks in vertical intelligent medicine deployment, including domain data scarcity, uneven data distribution, high annotation costs, limited generalization ability, black-box reasoning opacity, excessive computational overhead, privacy risks, and difficulties in adapting to rapidly updated clinical knowledge. To address these limitations, emerging small-scale models enable flexible horizontal expansion, offering lightweight, low-cost, and deployable complements to heavyweight LLMs. Beyond model-level optimization, reliable clinical adoption relies on effective human–AI synergy to constrain model hallucinations, standardize clinical reasoning, and ensure controllable and interpretable intelligent decision-making. This study presents a systematic intelligent medicine research paradigm integrating vertical LLM localization, horizontal small-model extension, and human-in-the-loop collaborative validation. We summarize practical solutions covering multi-source data standardization and secure sharing, domain knowledge graph-enhanced reasoning, specialized disease scenario deployment, and sustainable model iteration mechanisms. Combined with empirical case studies, this work further discusses the developmental pathways of horizontal small-model expansion and the future landscape of health data science and human–AI collaborative intelligent healthcare.

 

 

 

Prof. Kenta Nakai
Professor, Human Genome Center, The Institute of Medical Science, The University of Tokyo

Japan

Prof. Kenta Nakai is a distinguished Professor of Bioinformatics at the Human Genome Center, Institute of Medical Science, the University of Tokyo. Since earning his PhD from Kyoto University in 1992, he has spent decades advancing functional genomics and sequence analysis. His primary research focuses on computational biology, particularly developing bioinformatic tools and algorithms to analyze genomic sequences, gene regulation mechanisms, transcription factor binding sites, and protein subcellular localization. Among his notable contributions is PSORT, a widely used software tool that predicts protein subcellular localization based on sequence features. At the University of Tokyo, Prof. Nakai leads research initiatives that bridge computational methods with experimental biology to decode complex biological systems and human disease mechanisms. He has authored numerous peer-reviewed scientific articles and served in various leadership and academic roles within the global bioinformatics community, including leading the Japanese Society for Bioinformatics (JSBi) during fiscal year 2006. Through his research, teaching, and mentorship, Professor Nakai continues to play a prominent role in shaping computational genomics education and research both in Japan and internationally. He is the inaugural recipient of both the JSBi Prize (2020) and the APBioNet Research Innovation Award (2024), and was named a Fellow of the International Society for Computational Biology (ISCB) in the Class of 2026.

 

Speech Title: "Catching Crumbs from the Table": Can We Learn Experimentally Testable Knowledge from AI?

 

Abstract: Recent advancements in AI, such as AlphaFold, demonstrate unprecedented precision that outpaces human scientists. This shift mirrors Ted Chiang’s 2000 sci-fi novel, "Catching Crumbs from the Table," which depicts a future where "Metahuman" science surpasses human comprehension, reducing ordinary scientists to interpreters of the "crumbs" left behind. As AI continues to evolve, human researchers must proactively develop methodologies to extract and translate AI-learned patterns into comprehensible scientific knowledge. This talk introduces an ongoing initiative to extract knowledge from SpliceSelectNet (Miyachi and Nakai, NAR 2026), a novel hierarchical transformer-based predictor for RNA splice sites. Disrupted splice sites can alter splicing patterns and trigger genetic diseases. However, the precise rules governing splice-site selection remain elusive because introns are so large, making wet-lab exploration challenging. SpliceSelectNet overcomes this by predicting aberrant splicing patterns with relatively high accuracy. Furthermore, by analyzing attention patterns during in silico mutagenesis, we can identify critical cis-elements, such as exonic splicing enhancers, though their reliability needs further investigation. By systematically extracting these features, this project aims to elucidate splice-site selection rules and compile the extracted features/predicted changes in splicing patterns into a comprehensive database, ultimately advancing our understanding and treatment of various genetic diseases.

 

Prof. Minghui Li
Soochow University, China

 

Minghui Li is a Professor and PhD supervisor at the School of Basic Medical Sciences, Soochow University. Her research focuses on AI-enabled biomedicine, particularly protein mutation effect prediction and the identification of key cancer biomarkers. She received her Ph.D. in Physical Chemistry from Jilin University and conducted postdoctoral research at the University at Buffalo and the National Institutes of Health/National Center for Biotechnology Information (NIH/NCBI) in the United States. She serves as an Associate Editor of the Journal of Computational Biophysics and Chemistry and has published more than 30 papers in journals including PNAS, Cancer Research, Nucleic Acids Research, Communications Biology, and the Journal of Chemical Theory and Computation. Research group website: https://lilab.jysw.suda.edu.cn/

 

Speech Title: "An Interpretable Molecular Framework for Predicting Cancer Driver Missense Mutations"

 

Abstract: Missense mutations play a critical role in human disease, contributing to both inherited disorders and cancer. However, accurately predicting their functional impact—particularly for cancer driver mutations—remains a major challenge due to limited validated labels and the complex molecular basis of oncogenesis. Here, we systematically characterized over 120,000 missense variants across pathogenic, benign, driver, passenger, recurrent somatic, and common population classes, using a comprehensive set of mechanistically grounded molecular features. By assessing the statistical burden of variations, we demonstrated that these features effectively discriminate among diverse variant classes and reveal a consistent enrichment of functional sites, structural integrity, and biophysical changes in pathogenic and driver mutations. Building on these insights, we developed MutaPheno, an interpretable framework for predicting the functional consequences of missense mutations. The model integrates 34 molecular-level features, encompassing structural, functional, physicochemical, and contextual descriptors, using a random forest algorithm. Trained exclusively on pathogenic and benign variants, MutaPheno achieved strong accuracy in predicting cancer driver mutations, outperforming both cancer-specific and general pathogenicity tools, while also demonstrating superior robustness when tested on unseen proteins. Our findings highlight the shared mechanisms between pathogenic and driver mutations and emphasize the role of molecular features in improving variant interpretation. MutaPheno provides a transparent and generalizable tool that can facilitate driver discovery and the development of targeted therapies.

 

Prof. Rubita Sudirman
Universiti Teknologi Malaysia, Malaysia

 

Rubita Sudirman holds a Bachelor's (Hons) and Master's degree from the University of Tulsa, USA, and a Ph.D. in Electrical Engineering from Universiti Teknologi Malaysia (UTM). She is a professor and a certified professional engineer at Faculty of Electrical Engineering, UTM. Her research interests focus on the application of soft computing in biomedical engineering, particularly in speech processing, electroencephalography (EEG) and electrooculography (EOG) signal analysis, medical electronics, and rehabilitation engineering.

 

Speech Title: "Caffeine-Intake Variables based on Physiological Indicators for Cognitive Function using Response Surface Methodology"

 

Abstract: Caffeine is the most widely consumed psychoactive substance, yet how everyday intake shapes the physiological signals underlying cognitive function is rarely characterised outside the laboratory. A pilot low-cost pipeline, pairing a research-grade wearable (Shimmer3+ GSR) with a purpose-built web platform, within a Response Surface Methodology (RSM) design to model how two caffeine-intake variables, dose and time post-consumption, relate to physiological indicators (electrodermal activity and heart rate) and self-reported state affect. Nine healthy young adults completed counterbalanced within subject crossover of three everyday doses—decaffeinated control (~2 mg), one instant-coffee sachet (~70 mg) and two sachets (~140 mg)—with skin conductance and photoplethysmography-derived heart rate sampled at 128 Hz during a baseline and at 5, 30 and 55 minutes post-consumption. A repeated cognitive battery (Stroop, visual short-term memory, RSVP and serial subtraction) and visual-analogue mood scales were administered each window. A second-order mixed-effects response surface was applied to baseline corrected, dose-aligned features. Preliminary, unvalidated patterns suggested skin conductance may increase with dose, with heart rate stable and no adverse affective response; cognitive analyses are ongoing. The study demonstrates a feasible RSM-based physiological-monitoring pipeline and effect-size estimates to design a larger, age-stratified study.

 

Assist. Prof. Faez Iqbal Khan
Xi'an Jiaotong-Liverpool University, China

 

Accomplished researcher and educator in Biotechnology, Bioinformatics, and Computational Chemistry with expertise in molecular dynamics, protein engineering, and AI-based drug design. Recognized among the Stanford University Top 2% Scientists (2023). Skilled in multi-locational, research-led, and technology-enhanced transnational education with a proven record of student-centered teaching, interdisciplinary collaborations, and leadership in academic development at XJTLU and beyond.

Assist. Prof. Yitao Yang
The University of Tokyo, Japan

 

Dr. Yitao Yang is a researcher working at the interface of bioinformatics, artificial intelligence, and systems biology. His research focuses on computational methods for single-cell and spatial transcriptomics, with particular interest in using foundation models to learn biologically meaningful representations across heterogeneous datasets. He develops language-grounded approaches that connect molecular profiles with biological knowledge, enabling robust characterization of cell identity, state, and tissue context. His current work investigates virtual-cell modeling and AI-driven in silico perturbation to generate testable hypotheses about disease-associated gene programs and cell–cell interactions. By integrating multimodal genomic data with prior biological knowledge, he aims to improve mechanistic insight and support precision medicine.

 

Speech Title: "LingoCell: A Language-Grounded Foundation Model for Disentangling Cell Identity and Spatial Niches "

 

Abstract: Single-cell and spatial transcriptomics are reshaping our ability to characterize tissue organization, but cell identity, batch variation, and local niche context remain difficult to separate in a unified representation. In this talk, I will introduce LingoCell, a language-grounded foundation model designed to link cellular profiles with biologically meaningful semantic representations. LingoCell produces frozen embeddings that preserve cell-type structure across datasets, platforms, and tissues while retaining biological conservation without explicit batch-integration objectives. Beyond representation learning, the shared semantic space enables disease descriptions to define interpretable directions in cell space. By perturbing genes in silico and quantifying their effects along these disease axes, LingoCell identifies cell-type-specific gene programs associated with lung disease and tumor ecosystems. Closed-loop tests and independent atlas analyses support the biological relevance of the nominated programs. Together, these results illustrate how language-grounded models can connect robust cellular representations with hypothesis generation, offering a route toward interpretable virtual-cell systems for mechanism discovery in health and disease.

 

Prof. Yumei Li
Soochow University, China

 

Dr. Yumei Li is a professor at School of Basic Medicine, Soochow University. Her research focuses on transcriptomics and epigenomics, with particular interest in the development and application of bioinformatics approaches to study epigenetic regulation in both fundamental biological processes and disease pathogenesis. She has authored more than 20 publications in high-impact journals, including PNAS, Molecular Cell, and Genome Biology. Dr. Li is a member of the Omics Branch of the Chinese Stroke Association. She has been recognized as a Jiangsu Specially-Appointed Professor, and her work is supported by grants from the National Natural Science Foundation of China (NSFC) and the Natural Science Foundation of Jiangsu Province.

 

Speech Title: "Decoding multidimensional epigenetic information from cell-free DNA for liquid biopsy"

 

Abstract: Cell-free DNA (cfDNA) provides a minimally invasive window into tissue-specific molecular alterations, but fully exploiting its diagnostic potential requires computational approaches that capture, resolve, and integrate diverse epigenetic signals. We developed a series of complementary frameworks for multidimensional cfDNA analysis. MESA enables simultaneous profiling and integration of diverse epigenetic features from a single experimental assay. To investigate tissue-of-origin inference, we systematically benchmarked five methylation-based cfDNA deconvolution methods, identifying CelFEER and UXM as consistently strong-performing approaches and demonstrating the importance of sequencing depth and reference atlas comprehensiveness. We further developed cfDecon, a deep-learning framework for read-level cfDNA deconvolution that estimates cell-type proportions while reconstructing condition-aware cell-type-specific methylation profiles. For disease detection, cfMIND leverages read-level methylation patterns to preserve heterogeneous disease-associated signals that may be obscured by conventional region-level averaging, enabling robust detection across multiple sequencing technologies and disease settings. Finally, cfDNAanalyzer provides a comprehensive computational toolkit for extracting, processing, integrating, and modeling diverse cfDNA features, including fragmentation, nucleosome, copy-number, and methylation-associated signals. Together, these approaches provide an integrated computational framework for decoding multidimensional cfDNA information and advancing liquid biopsy applications.

 

 

Previous Speakers

 

Prof. Yuan-Ting Zhang City University of Hongkong	Prof. Alexander Suvoror Institute of Experimental Medicine, St. Petersburg	Prof. David Zhang The Chinese University of Hongkong, China (Shenzhen)
Prof.Tun-Wen Pai National Taipei University of Technology	Prof. Dong-Qing Wei Shanghai Jiaotong University	Prof. TSUI Kwok-Wing Stephen The Chinese University of Hongkong
Prof. Cathy Wu University of Delaware	Prof. Xuegong Zhang Tsinghua University	Prof. Yi Pan Chinese Academy of Sciences
Prof. Bairong Shen Sichuan University	Prof. Wing-Kin Sung The Chinese University of Hongkong, and Hongkong Genome Institute	Prof. Chanchal Mitra University of Hyderabad
Assoc. Prof. Jie Zheng ShanghaiTech University	Prof. Peiyu Zhang Henan University	Prof. Zheng Zhou Chinese Academy of Sciences
Prof. Le Zhang Sichuan University	Prof. Fei Guo Central South University	Prof. Bin Liu Beijing Institute of Technology
Mr. Xiaoqiang Li China National GeneBank DataBase	Prof. Guan Ning Lin Shanghai Jiao Tong University	Assoc. Prof. Hon-Cheong So The Chinese University of Hongkong
Prof. Limsoon Wong (ACM Fellow) National University of Singapore	Prof. Bing Zhang Shanghai Jiao Tong University	Prof. An-Yuan Guo West China Hospital, Sichuan University
Prof. Pui-Chi Gigi Lo City University of Xiamen	Asst. Prof. Mengsha Tong Xiamen University	Prof. Jose Nacher Toho University
Prof. Rongshan Yu(SMIEEE, FIET) Vice Director, National Institute for Data Science in Health and Medicine, Xiamen University	Prof. Yasukazu Nakamura (H-index: 66) National Institute of Genetics	Prof. Xiaopei Shen Fujian Medical University
Prof. Yumei Li Soochow University	Assoc. Prof. Balachandran Manavalan Sungkyunkwan University	Dr. Jingjing Liu Hong Kong University of Science and Technology
Dr. Jin Wang Capital Medical University	Dr. Yaling Zhu Anhui Medical University	Assist. Prof. Yinran Chen Xiamen University
Assist. Prof. Fei Qi Xiamen University