Wang Gaowei, Su Hang, Yu Helin, Yuan Ruoshi, Zhu Xiaomei, Ao Ping
Ministry of Education Key Laboratory of Systems Biomedicine, Shanghai Center for Systems Biomedicine, Shanghai Jiao Tong University, Shanghai, China.
GenMath, Seattle, WA, USA.
J R Soc Interface. 2016 Feb;13(115):20151115. doi: 10.1098/rsif.2015.1115.
Cancers have been typically characterized by genetic mutations. Patterns of such mutations have traditionally been analysed by posteriori statistical association approaches. One may ponder the possibility of a priori determination of any mutation regularity. Here by exploring biological processes implied in a mechanistic theory recently developed (the endogenous molecular-cellular network theory), we found that the features of genetic mutations in cancers may be predicted without any prior knowledge of mutation propensities. With hepatocellular carcinoma (HCC) as an example, we found that the normal hepatocyte and cancerous hepatocyte can be represented by robust stable states of one single endogenous network. These stable states, specified by distinct patterns of expressions or activities of proteins in the network, provide means to directly identify a set of most probable genetic mutations and their effects in HCC. As the key proteins and main interactions in the network are conserved through cell types in an organism, similar mutational features may also be found in other cancers. This analysis yielded straightforward and testable predictions on accumulated and preferred mutation spectra in normal tissue. The validation of predicted cancer state mutation patterns demonstrates the usefulness and potential of a causal dynamical framework to understand and predict genetic mutations in cancer.
癌症通常以基因突变作为特征。此类突变模式传统上是通过事后统计关联方法进行分析的。人们可能会思考先验确定任何突变规律的可能性。在此，通过探索最近发展的一种机制理论（内源性分子 - 细胞网络理论）中所蕴含的生物学过程，我们发现无需任何关于突变倾向的先验知识，就可以预测癌症中基因突变的特征。以肝细胞癌（HCC）为例，我们发现正常肝细胞和癌性肝细胞可以由一个单一内源性网络的稳健稳定状态来表示。这些由网络中蛋白质表达或活性的不同模式所确定的稳定状态，提供了直接识别一组最可能的基因突变及其在HCC中的作用的方法。由于网络中的关键蛋白质和主要相互作用在生物体的细胞类型中是保守的，所以在其他癌症中可能也会发现类似的突变特征。该分析对正常组织中累积和偏好的突变谱产生了直接且可检验的预测。对预测的癌症状态突变模式的验证证明了因果动力学框架在理解和预测癌症基因突变方面的有用性和潜力。
