arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2501.19178cs.LG
收藏

没有基础就没有基础模型——为何半机制模型对调控生物学至关重要

No Foundations without Foundations -- Why semi-mechanistic models are essential for regulatory biology

University of Cambridge(剑桥大学);Massachusetts Institute of Technology(麻省理工学院);Broad Institute of MIT and Harvard(麻省理工学院与哈佛大学博德研究所)

Luka Kovačević, Thomas Gaudelet, James Opzoomer, Hagen Triendl, John Whittaker, Caroline Uhler, Lindsay Edwards, Jake P. Taylor-King

AI总结:

针对深度学习在预测基因表达等调控生物学上效果不佳的问题,提出一种整合机制见解与实验设计的半机制框架,通过改进损失函数和误差分析提升预测性能并优化分批策略。

摘要:

尽管付出了大量努力,但深度学习尚未在阐明调控生物学方面产生变革性影响,特别是在预测基因表达谱的领域。我们认为,如果不结合将机制性见解与规范性实验设计相整合的框架,调控生物学真正的“基础模型”将仍然遥不可及。我们提出了一个自下而上的半机制框架,该框架统一了体外和体内 CRISPR 筛选中基于扰动的实验设计,并考虑了分化和非分化细胞系统。通过揭示已发表的机器学习方法中先前未被认识到的假设,我们的方法澄清了其与变分自编码器和结构因果模型等流行技术的联系。在实践中,该框架提出了一种改进的损失函数,我们证明它可以提高预测性能,并进一步提出了一种为分批策略提供信息的误差分析。最终,由于细胞调控产生于大量未知的分子成分之间的无数相互作用,我们认为仅靠结构生物学无法实现系统层面的理解。相反,我们主张真正的进步需要对实验如何捕获生物现象、数据如何生成,以及这些过程如何反映在更忠实的建模架构中采取第一性原理的视角。

英文摘要:

Despite substantial efforts, deep learning has not yet delivered a transformative impact on elucidating regulatory biology, particularly in the realm of predicting gene expression profiles. Here, we argue that genuine "foundation models" of regulatory biology will remain out of reach unless guided by frameworks that integrate mechanistic insight with principled experimental design. We present one such ground-up, semi-mechanistic framework that unifies perturbation-based experimental designs across both in vitro and in vivo CRISPR screens, accounting for differentiating and non-differentiating cellular systems. By revealing previously unrecognised assumptions in published machine learning methods, our approach clarifies links with popular techniques such as variational autoencoders and structural causal models. In practice, this framework suggests a modified loss function that we demonstrate can improve predictive performance, and further suggests an error analysis that informs batching strategies. Ultimately, since cellular regulation emerges from innumerable interactions amongst largely uncharted molecular components, we contend that systems-level understanding cannot be achieved through structural biology alone. Instead, we argue that real progress will require a first-principles perspective on how experiments capture biological phenomena, how data are generated, and how these processes can be reflected in more faithful modelling architectures.

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。

↑