Annual Report 2023
Laboratory of Computational Life Science
Yasuhiro Kojima, Naoko Takagi
Introduction
Recent advancements in single-cell and spatial omics analysis have significantly increased the complexity of omics data, particularly transcriptomics. The Computational Life Science Unit is developing computational methodologies to efficiently extract knowledge from these comprehensive gene expression profiles for single cells and spatial niches within tissues.
The Team and What We Do
We are developing computational methodologies by combining mathematical models and deep learning. We have developed a method for extracting cell-cell interaction information from spatial omics data, and we are also expanding collaborative research utilizing this method. Additionally, by combining this with the cell state dynamics estimation method developed in the previous year, we have established a method for estimating cell state dynamics dependent on the microenvironment.
Research Activities
1. Development of a deep generative model for co-localization analysis at the single-cell level
While spatial omics analysis provides comprehensive gene expression profiles with spatial context, the most widely adopted methods have not achieved single-cell resolution. Therefore, in this study, we developed DeepCOLOR, a technique that estimates the spatial distribution of single-cell transcriptomes within tissues by integrating with single-cell transcriptomics. This enabled the construction of an analytical framework that allows for the identification of cell populations defined by co-localization through the establishment of single-cell level co-localization networks.
2. Elucidation of immune tolerance mechanisms through co-localization analysis in early colorectal cancer
By applying the DeepCOLOR, we estimated co-localization networks in early colorectal cancer. In this project, we discovered a cancer cell population specifically present in the adenoma region of early cancer that highly expresses the extracellular signaling molecule MDK. We also demonstrated that this is involved in the activation of Tregs in adenomas.
3. Integrated analysis of single-cell transcriptomics and live imaging in zebrafish embryonic development
What molecular foundations support the dynamic cellular dynamics during embryonic development? In this study, we applied pseudotime estimation and change point detection techniques to live imaging data and single-cell transcriptome data during the tail elongation process in zebrafish embryonic development. This allowed us to identify regions that consistently change in both cellular dynamics and molecular profiles.
Education
We provide research guidance to several students, primarily from the Department of Systems Biology at Nagoya University. Approximately three students have already submitted their papers for publication. Additionally, we have started research with a doctoral student who will be joining us as a JSPS Postdoctoral Fellow in the next academic year.
Future Prospects
Current advancements in spatial transcriptomics are leading to increased resolution, enabling analysis at subcellular levels. Moving forward, we will develop techniques for representation learning of microenvironments composed of diverse cell types and for estimating their dynamics.
List of papers published in 2023
Journal
1. Kojima Y, Mii S, Hayashi S, Hirose H, Ishikawa M, Akiyama M, Enomoto A, Shimamura T. Single-cell colocalization analysis using a deep generative model. Cell systems, 15:180-192.e7, 2024
2. Koseki J, Hayashi S, Kojima Y, Hirose H, Shimamura T. Topological data analysis of protein structure and inter/intra-molecular interaction changes attributable to amino acid mutations. Computational and structural biotechnology journal, 21:2950-2959, 2023
3. Genuth MA, Kojima Y, Jülich D, Kiryu H, Holley SA. Automated time-lapse data segmentation reveals in vivo cell state dynamics. Science advances, 9:eadf1814, 2023
4. Nakahara R, Aki S, Sugaya M, Hirose H, Kato M, Maeda K, Sakamoto DM, Kojima Y, Nishida M, Ando R, Muramatsu M, Pan M, Tsuchida R, Matsumura Y, Yanai H, Takano H, Yao R, Sando S, Shibuya M, Sakai J, Kodama T, Kidoya H, Shimamura T, Osawa T. Hypoxia activates SREBP2 through Golgi disassembly in bone marrow-derived monocytes for enhanced tumor growth. The EMBO journal, 42:e114032, 2023
5. Ishikawa M, Sugino S, Masuda Y, Tarumoto Y, Seto Y, Taniyama N, Wagai F, Yamauchi Y, Kojima Y, Kiryu H, Yusa K, Eiraku M, Mochizuki A. RENGE infers gene regulatory networks using time-series single-cell RNA-seq data with CRISPR perturbations. Communications biology, 6:1290, 2023