Scientists traditionally examine tissues by analyzing gene expression levels in individual cells using a technique known as spatial transcriptomics (ST). Researchers gain information about the spatial organization and function of cells by measuring the amount of RNA at specific locations within a tissue. ST spatial transcriptomics technologies have been instrumental in unraveling the mysteries of mRNA expression in single cells while maintaining their spatial coordinates. However, challenges arise when multiple tissue slices need to be analyzed together and the size of the spots on ST slices makes resolution difficult.
To overcome these limitations, a group of researchers led by Prof. Qu Kun of the University of Science and technology of the Chinese Academy of Sciences has created a solution called Spatial Architecture Characterization through Deep Learning (SPACEL). This toolset has three modules (Spoint, Splane and Scube) that combine to automatically create a 3D tissue panorama.
The first module, Sprint, addresses the task of cell type deconvolution. It predicts the spatial distribution of cell types using a combination of simulated pseudopoints, neural network modeling, and statistical retrieval of expression profiles. This makes the predictions accurate and powerful. The second module, Splane, uses a graph convolutional network (GCN) approach and an adversarial learning algorithm to identify special domains by jointly analyzing multiple ST slices. Splane uses adversarial training to eliminate batch effects in multiple sectors and uses cell type composition as input. Splane stands out for its innovative method of efficient identification of spatial domains. The third module, Scube, automates slice alignment and builds a stacked 3D architecture of the tissue. This is crucial to overcome the challenges posed by the limitations of ST experimental techniques, allowing for a comprehensive understanding of the three-dimensional structure of the tissue.
The researchers applied SPACEL to 11 ST data sets with a total of 156 slices and used technologies such as 10X Visium, STARmap, MERFISH, Stereo-seq, and Spatial Transcriptomics. The researchers emphasize that SPACEL outperformed previous techniques in three fundamental analytical tasks: cell type distribution prediction, spatial domain identification, and three-dimensional tissue reconstruction.
Furthermore, SPACEL demonstrated its superiority in cell type deconvolution, spatial domain identification, and 3D alignment against 19 state-of-the-art techniques on simulated and real ST datasets, with its superior performance over previous techniques and a simplified approach to understanding accurately ST data.
In conclusion, the introduction of SPACEL is an important step in spatial transcriptomics. Its three modules provide researchers with a powerful tool to overcome the challenges associated with joint analysis of multiple ST slices, enabling accurate cell type predictions, effective spatial domain identification, and precise 3D tissue alignment. This tool enables accurate 3D tissue alignment, cell type predictions, and efficient spatial domain identification.
Review the Paper. All credit for this research goes to the researchers of this project. Also, don't forget to join. our SubReddit of more than 35,000 ml, 41k+ Facebook community, Discord channel, LinkedIn Grabove, Twitterand Electronic newsletterwhere we share the latest news on ai research, interesting ai projects and more.
If you like our work, you'll love our newsletter.
Rachit Ranjan is a consulting intern at MarktechPost. He is currently pursuing his B.tech from the Indian Institute of technology (IIT), Patna. He is actively shaping his career in the field of artificial intelligence and data science and is passionate and dedicated to exploring these fields.
<!– ai CONTENT END 2 –>
