Computational approaches for the analyses of spatial profiling technologies

在肿瘤生态系统中，不同细胞群之间的位置和空间相互作用决定了肿瘤的进展和治疗结果。然而，传统的基因组学方法，如批量测序和单细胞分析，不能捕捉关键的空间信息。近年来，蛋白质组学和转录组学分析的空间技术都取得了快速进展。这些技术产生了新的数据类型，其分辨率可能低于当前数据，但包含空间动态模式。因此，我计划开发生物信息学方法，使空间蛋白质组学和转录组学数据的知识挖掘。我们的主要重点是开发一个自动注释框架，从空间转录组数据中识别癌症，基质和免疫区域。同时，我们的方法还将分解免疫细胞的详细谱系，并分析空间相互作用的富集。

In a tumor ecosystem, the location and spatial interactions among diverse cell populations are determining the tumor progression and therapy outcome. However, conventional genomics approaches, such as bulk sequencing and single-cell profiling, cannot capture critical spatial information. Recent years have seen the rapid progress of spatial technologies for both proteomics and transcriptomics profiling. These technologies generate new data types, which may have a lower resolution than current data, but contains spatial dynamics patterns. Therefore, I plan to develop bioinformatics methods to enable knowledge mining for spatial proteomics and transcriptomics data. Our primary focus is to develop an automatic annotation framework to identify cancer, stroma, and immune regions from the spatial transcriptomic data. Meanwhile, our method will also decompose the detailed lineages of immune cells and analyze the spatial interaction enrichments.