arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.17532q-bio.GNcs.AI
收藏

单细胞基础模型解释性系统评估揭示注意力捕捉共表达而非独特调控信号

Systematic Evaluation of Single-Cell Foundation Model Interpretability Reveals Attention Captures Co-Expression Rather Than Unique Regulatory Signal

Department of Computer Science University of Tübingen(计算机科学系图宾根大学)

Ihor Kendiukhov

AI总结:

单细胞基础模型中注意力机制捕捉共表达而非独特调控信号,CSSI提升GRN恢复效率

摘要:

我们提出了一种系统评估框架-三十七项分析,153项统计检验,四种细胞类型,两种扰动模式-用于评估单细胞基础模型中的机制可解释性。将此框架应用于scGPT和Geneformer,我们发现注意力模式编码了结构化的生物信息,具有层特定的组织-早期层中的蛋白质-蛋白质相互作用,后期层中的转录调控-但这种结构对扰动预测没有增量价值:平凡的基因层面基准优于注意力和相关边(AUROC 0.81-0.88 vs 0.70),成对边分数增加零预测贡献,因果消融调控头不产生退化。这些发现从K562到RPE1细胞普遍有效;注意力-相关关系是情境依赖的,但基因层面主导是普遍的。Cell-State Stratified Interpretability (CSSI) 解决了注意力特定的缩放失败,将GRN恢复提高到1.85倍。该框架为该领域建立了可重用的质量控制标准。

英文摘要:

We present a systematic evaluation framework - thirty-seven analyses, 153 statistical tests, four cell types, two perturbation modalities - for assessing mechanistic interpretability in single-cell foundation models. Applying this framework to scGPT and Geneformer, we find that attention patterns encode structured biological information with layer-specific organisation - protein-protein interactions in early layers, transcriptional regulation in late layers - but this structure provides no incremental value for perturbation prediction: trivial gene-level baselines outperform both attention and correlation edges (AUROC 0.81-0.88 versus 0.70), pairwise edge scores add zero predictive contribution, and causal ablation of regulatory heads produces no degradation. These findings generalise from K562 to RPE1 cells; the attention-correlation relationship is context-dependent, but gene-level dominance is universal. Cell-State Stratified Interpretability (CSSI) addresses an attention-specific scaling failure, improving GRN recovery up to 1.85x. The framework establishes reusable quality-control standards for the field.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑