Teschendorff Andrew E, Sollich Peter, Kuehn Reimer
CAS-MPG Partner Institute for Computational Biology, Chinese Academy of Sciences, Shanghai Institute for Biological Sciences, 320 Yue Yang Road, Shanghai 200031, China; Statistical Cancer Genomics, Paul O'Gorman Building, UCL Cancer Institute, University College London, London WC1E 6BT, UK.
Department of Mathematics, King's College London, London WC2R 2LS, UK.
Methods. 2014 Jun 1;67(3):282-93. doi: 10.1016/j.ymeth.2014.03.013. Epub 2014 Mar 24.
A key challenge in systems biology is the elucidation of the underlying principles, or fundamental laws, which determine the cellular phenotype. Understanding how these fundamental principles are altered in diseases like cancer is important for translating basic scientific knowledge into clinical advances. While significant progress is being made, with the identification of novel drug targets and treatments by means of systems biological methods, our fundamental systems level understanding of why certain treatments succeed and others fail is still lacking. We here advocate a novel methodological framework for systems analysis and interpretation of molecular omic data, which is based on statistical mechanical principles. Specifically, we propose the notion of cellular signalling entropy (or uncertainty), as a novel means of analysing and interpreting omic data, and more fundamentally, as a means of elucidating systems-level principles underlying basic biology and disease. We describe the power of signalling entropy to discriminate cells according to differentiation potential and cancer status. We further argue the case for an empirical cellular entropy-robustness correlation theorem and demonstrate its existence in cancer cell line drug sensitivity data. Specifically, we find that high signalling entropy correlates with drug resistance and further describe how entropy could be used to identify the achilles heels of cancer cells. In summary, signalling entropy is a deep and powerful concept, based on rigorous statistical mechanical principles, which, with improved data quality and coverage, will allow a much deeper understanding of the systems biological principles underlying normal and disease physiology.
系统生物学中的一个关键挑战是阐明决定细胞表型的潜在原理或基本规律。了解这些基本原理在癌症等疾病中是如何改变的，对于将基础科学知识转化为临床进展至关重要。虽然通过系统生物学方法在识别新型药物靶点和治疗方法方面取得了重大进展，但我们在系统层面上对某些治疗成功而其他治疗失败的根本原因仍缺乏了解。我们在此倡导一种基于统计力学原理的用于分子组学数据系统分析和解释的新型方法框架。具体而言，我们提出细胞信号熵（或不确定性）的概念，作为分析和解释组学数据的一种新方法，更根本地说，作为阐明基础生物学和疾病背后系统层面原理的一种方法。我们描述了信号熵根据分化潜能和癌症状态区分细胞的能力。我们进一步论证了经验性细胞熵 - 稳健性相关定理的情况，并在癌细胞系药物敏感性数据中证明了它的存在。具体而言，我们发现高信号熵与耐药性相关，并进一步描述了熵如何用于识别癌细胞的致命弱点。总之，信号熵是一个基于严格统计力学原理的深刻而强大的概念，随着数据质量和覆盖范围的提高，将使我们能够更深入地理解正常和疾病生理学背后的系统生物学原理。
