Fan Xiaohui's research group at the School of Pharmacy of Zhejiang University reported in Nature Communications that the first ...

In recent years, single-cell sequencing (scRNA-seq) and spatial transcriptome sequencing (Spatially resolved transcriptomics) technologies have revolutionized the way people understand complex tissues, and continue to promote life and health research into the era of single-cell and spatial omics
。 However, current single-cell sequencing and spatial transcriptome technologies are not only time-consuming, labor-intensive, but also extremely expensive, which severely limits their popularization
.
At the same time, traditional mixed sequencing (Bulk RNA-seq) has been a mainstream tool for biomedical research, especially TCGA, ENCODE, and ICGC and other multi-scale projects have been carried out successively to accumulate massive data resources
for related research.
Therefore, how to make full use of the original data resources to obtain the spatial gene expression profile of single cell resolution from the Bulk data is a major technical challenge
in systems biology research.

On October 30, 2022, the team of Professor Fan Xiaohui from the School of Pharmacy of Zhejiang University and the Yangtze River Delta Smart Oasis Innovation Center, together with the team of Professor Chen Huajun of the School of Computer Science and Technology and the team of Professor Gao Yue of the Academy of Military Medical Sciences, are internationally renowned journal Nature A research paper titled "De novo analysis of bulk RNA-seq data at spatially resolved single-cell resolution" was published in Communications.

This study proposes a spatial deconvolution algorithm, Bulk2Space, which uses deep learning frameworks such as β-VAE to realize Bulking for the first time Transcriptome reconstitution to single-cell spatial resolution
.

The research team used Bulk2Space to reveal the differences in gene expression of B lymphocytes in different tumor regions and the spatial transcriptional heterogeneity of
tissues during the transformation of inflammatory cancer.
In this study, Bulk2Space was also applied to the spatial deconvolution analysis
of Bulking data in different brain regions of mouse brain tissue.
The relevant data are all collected from Spatial-seq, a high-throughput multi-sample spatial sequencing technology based on laser capture microdissection (LCM) independently developed by Professor Fan Xiaohui's team
。 Studies have shown that Bulk2Space can not only reconstruct the hierarchical spatial structure of the mouse cortex, but also re-annotate
unknown cell types in the hypothalamus.
A series of studies have confirmed that Bulk2Space not only shows excellent performance in simulation and real data sets, but also successfully applies to a variety of biology and disease scenarios, with a wide range of application prospects
.
The algorithm is now open sourced to the GitHub platform (https://github.
com/ZJUFanLab/bulk2space).

Schematic diagram of the Bulk2Space algorithm principle

Bulk2Space was used to spatially deconvolute and annotate the mouse hypothalamic bulk transcriptome obtained by Spatial-seq technology

The first authors of this paper are Liao Jie, postdoctoral fellow, Qian Jingyang, doctoral student of Zhejiang University School of Pharmacy, and Fang Yin, Chen Zhuo and Zhuang Xiang, doctoral students of School of Computer Science and Technology, Zhejiang
University.
The corresponding authors are Professor Fan Xiaohui, School of Pharmacy, Zhejiang University, Professor Chen Huajun, School of Computer Science and Technology, Zhejiang University, and Professor
Gao Yue, Institute of Radiation and Radiation Medicine, Academy of Military Medical Sciences.
This research was supported by the National Natural Science Foundation of China, the National Multidisciplinary Innovation Team of Traditional Chinese Medicine, and supported
by Alibaba Cloud.

Original link: