Novae: a graph-based foundation model for spatial transcriptomics data
7.4
来源：
Nature
关键字：
deep learning brain science
发布时间：
2025-12-11 03:40
摘要：
Novae is a novel graph-based foundation model designed for spatial transcriptomics, significantly enhancing the analysis of gene expression in various tissues. Trained on a dataset of nearly 30 million cells, it allows for zero-shot domain inference and effectively corrects batch effects. The model supports multiple downstream tasks, making it a versatile tool for biomedical research, particularly in understanding tissue microenvironments and disease progression.
原文：
查看原文
价值分投票
评分标准
新闻价值分采用0-10分制，综合考虑新闻的真实性、重要性、时效性、影响力等多个维度。
评分越高，表示该新闻的价值越大，越值得关注。
价值维度分析
domain_focus
1.0分+重点关注领域符合度
business_impact
0.5分+商业影响力
scientific_rigor
1.5分+数据支撑的科学性
timeliness_innovation
1.5分+时效性与创新性
investment_perspective
2.5分+BOCG投资视角
market_value_relevance
1.0分+市场价值相关性
team_institution_background
0.5分+团队与机构背景
technical_barrier_competition
0.4分+技术壁垒与竞争格局
关键证据
Novae is trained on a large dataset of nearly 30 million cells across 18 tissues.
The model enables zero-shot domain inference across multiple gene panels, tissues, and technologies.
It provides a robust tool for advancing spatial transcriptomics and its applications in biomedical research.
真实性检查
否
AI评分总结
Novae is a novel graph-based foundation model designed for spatial transcriptomics, significantly enhancing the analysis of gene expression in various tissues. Trained on a dataset of nearly 30 million cells, it allows for zero-shot domain inference and effectively corrects batch effects. The model supports multiple downstream tasks, making it a versatile tool for biomedical research, particularly in understanding tissue microenvironments and disease progression.