Skip to content
1183-Cellist: Cell Segmentation for Spatial Transcriptomics episode artwork

Jun 25, 2026 17 min

1183-Cellist: Cell Segmentation for Spatial Transcriptomics

from Paper Talk

This technical report introduces Cellist, a multi-modal computational method designed to improve cell segmentation in high-resolution spatial transcriptomics. By integrating visual information from staining images with molecular expression data, the tool addresses the common challenge of transcript diffusion, which often leads to blurred boundaries and inaccurate cell identification. Research demonstrates that Cellist consistently outperforms existing methods across diverse platforms, such as Stereo-seq and 10x Xenium, by maintaining superior transcriptomic integrity and biological specificity. Its high computational efficiency and scalability make it suitable for analyzing large datasets with millions of spatial spots. The authors successfully applied the tool to lung cancer samples, uncovering the spatial organization of tumor clones and identifying specific immune cell subtypes linked to immunotherapy responses. Ultimately, the study highlights how accurate cell-level analysis can reveal intricate tissue architectures and complex intercellular interactions within the tumor microenvironment. References: Sun D, Zhang L, Han T, et al. Accurate, scalable and cross-platform cell identification for high-resolution spatial transcriptomics[J]. Nature Genetics, 2026: 1-12. 前往小宇宙评论区与主播互动

Episode details come from the podcast's RSS feed.

Embed this episode

Ready to play

1183-Cellist: Cell Segmentation for Spatial Transcriptomics

0:00 17:55

No transcript yet.

Requests help us decide what to transcribe next.

Frequently asked questions

How long is this episode of Paper Talk?

This episode is 17 minutes long.

When was this Paper Talk episode published?

This episode was published on June 25, 2026.

Can I download this Paper Talk episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.