Lecture review

Lecture 1: “SuperSCC: single cell hierarchical clustering reveals cross-tissue conserved gene modules”

Introduction to the speaker

Tang Feng, a postdoctoral fellow at Guangzhou National Laboratory and a researcher at the cooperative instructor Tian Luyi. He graduated from the University of Adelaide, Australia with a Ph.D. majoring in biomedical science at the University of Adelaide, Australia, focusing on data analysis and algorithm development of single-cell RNA sequencing and spatial transcriptomics. He is mainly engaged in innovative algorithm design and biomedical applications, committed to revealing the molecular mechanisms of complex diseases, and providing technical support for disease mechanism analysis and precision medicine. Currently, 5 SCI papers have been published as the first author.

Main content

Dr. Tang Feng systematically elaborates on the cutting-edge challenges in single-cell RNA sequencing (scRNA-seq) data analysis and its innovative research in the analysis of dynamic changes in cell state. In response to the challenges of batch effect interference in single-cell RNA sequencing data analysis, difficulty in cell identity analysis, complex identification of conservative gene modules across data sets, and the traditional analysis process often ignores the dynamic changes in cells in pathological states and is difficult to reveal the limitations of key cellular procedures for disease evolution, Dr. Tang Feng reported his latest developed SuperSCC data analysis framework: by integrating supersed feature selection, graph clustering and hierarchical merging strategies, it is possible to accurately identify cell types and states without data integration, and to efficiently discover cross-cohort gene modules. Research shows that in pathological scenarios such as lung cancer and immune microenvironment, SuperSCC can successfully explore cross-tissue conservative gene modules, accurately identify rare cell status, and effectively predict immunotherapy responses. ； Its feature of data integration without data integration not only simplifies the analysis process, but also breaks through the bottlenecks of traditional methods when dealing with complex data sets. This framework not only provides new ideas for the construction of complex tissue cell maps, but also reveals the dynamic changes of key cellular programs in disease evolution, and provides more biologically significant analytical tools for disease research and the development of precision medicine.

Lecture 2: “Advancing sequencing-based spatial transcriptomics with a comprehensive benchmarking study”

Introduction to the speaker

You Yue, a postdoctoral fellow at Guangzhou National Laboratory and a researcher at the cooperative instructor Tian Luyi. He graduated from the University of Melbourne, Australia with a Ph.D. majoring in biological sciences at the University of Melbourne, Australia. His research directions are single-cell multiomics algorithm development and spatial omics data analysis. He presided over 1 project for postdoctoral funding (on-site) of the National Natural Science Foundation of China and 1 project in each. As the first or corresponding author in Nature Methods, Signal Transduction and Targeted Therapy, Genome Biology and other top international journals have published many papers.

Main content

Dr. You Yue mainly introduced the cutting-edge progress in the standardization research of spatial transcriptomics (sST) technology and its breakthrough application in solving the problem of differentiation of technology platforms. In view of the huge differences in resolution, capture efficiency and diffusion characteristics of different technical platforms in the field of spatial transcriptomics, and the lack of unified benchmarks, the "bottleneck" link in the selection and algorithm development of methods, Dr. You Yue published it with his team in "Nature The first systematic study of Methods is the core, focusing on introducing key innovative achievements: by constructing a reference tissue library with a clear histological structure covering the mouse brain, kidney, small intestine, etc., 11 mainstream sST technologies were evaluated horizontally in 35 groups of experiments, and for the first time "molecular diffusion" was quantified as a core variable that affects effective resolution. At the same time, its research team developed a reproducible slice and permeability optimization protocol and an open database genographix.com, successfully realizing the full process standardization from experimental design, parameter calibration to calculation and evaluation ； The unified benchmark it established not only provides a reliable basis for the accurate identification of low-abundance rare cell states, but also effectively breaks through the technical barriers of cross-platform data integration. This work not only reveals the key influence mechanism of molecular diffusion on the performance of spatial transcriptomics technology, but also provides a "gold standard" and one-stop tool for the subsequent iteration of spatial multiomics technology, demonstrating the core value of standardized research in promoting spatial transcriptomics from technological exploration to widespread application.