Detecting anomalous anatomic regions in spatial transcriptomics with STANDS
Lists
Only you see this
Abstract
Detection and Dissection of Anomalous Tissue Domains (DDATD) from multi-sample spatial transcriptomics (ST) data provides unprecedented opportunities to characterize anomalous tissue domains (ATDs), revealing both population-level and individual-specific pathogenic factors for understanding pathogenic heterogeneities behind diseases. However, no current methods can perform de novo DDATD from ST data, especially in the multi-sample context. Here, we introduce STANDS, an innovative framework based on Generative Adversarial Networks which integrates three core tasks in multi-sample DDATD: detecting, aligning, and subtyping ATDs. STANDS incorporates multimodal-learning, transfer-learning, and style-transfer techniques to effectively address major challenges in multi-sample DDATD, including complications caused by unalignable ATDs, under-utilization of multimodal information, and scarcity of normal ST datasets necessary for comparative analysis. Extensive benchmarks from diverse datasets demonstrate STAND’s superiority in identifying both common and individual-specific ATDs and further dissecting them into biologically distinct subdomains. STANDS also provides clues to developing ATDs visually indistinguishable from surrounding normal tissues. The authors introduce STANDS, a GAN-based framework that integrates three core tasks for the multi-sample detection and dissection of anomalous tissue domains from spatial transcriptomics data, revealing pathogenic heterogeneity behind diseases.
BibTeX
@inproceedings{xu2024detecting,
author = {Kaichen Xu and Yan Lu and Suyang Hou and Kainan Liu and Yihang Du and Mengqian Huang and Hao Feng and Hao Wu and Xiaobo Sun},
title = {{Detecting anomalous anatomic regions in spatial transcriptomics with STANDS}},
booktitle = {Nature Communications},
year = {2024},
url = {https://openreview.net/forum?id=aDd33rkh9f},
}Loading related content...