通过相对熵得分诊断个体状态

Diagnosing the State of Individuals by Relative Entropy Score

摘要: 复杂疾病的发展需要经历三种状态：正常状态、前疾病状态和疾病状态。此外，生物网络的正常状态与疾病状态之间存在着很大的差异。如果能获得网络在不同状态下的特征，便可以达到疾病预警的目的。在本研究中，我们提出一个算法学习网络在不同状态下的特征，从而判断个体状态。我们将该方法应用到真实数据集肺鳞状细胞癌，从而验证了该方法的有效性。

Abstract: The development of complex disease can be divided as three states as follows: a normal state, a pre-disease state, and a disease state. Furthermore, there are a lot of significant differences be-tween the network of the normal state and the disease state. If we can obtain the features of the network in normal and the disease state, we achieve the purpose of disease early warning. In this study, we proposed an algorithm to learn the features of different states, and thus distinguish dif-ferent states. We certificate the effectiveness of the method by applying this method to the data set Lung squamous cell carcinoma.