microRNA（简称miRNA）是一类重要的调控小RNA。异常的miRNA调控网络与癌症发生发展的各个阶段均有密切关系。在癌症里发生异常调控的miRNA被称为“癌症microRNA”（onco-microRNA，简称oncomiR），并与其异常调控的靶基因组成oncomiR调控网络。目前，已知调控机制的oncomiR较少，其调控网络的数学模型仍不明确。我们的研究结果以及相关生物学实验证据均表明miRNA调控呈现多尺度的特性：有的miRNA有数百甚至上千个靶基因，广泛调控各种生物过程；而有的仅有几十个靶基因，特异性调控个别生物过程。而癌症这类复杂疾病也涉及到大量的、不同尺度的生物过程的异常调控。因此，为了开展oncomiR调控的系统研究，我们计划首先建立基于多尺度调控特性的oncomiR调控网络的数学模型，然后结合大规模癌症临床样本的表达数据对oncomiR调控网络进行系统的识别与分析。

MicroRNAs (miRNAs) are a class of important regulatory small RNAs. The dys-regulations of miRNA regulatory networks are closely associated with multiple stages of cancer initiation and progression. The dys-regulated miRNAs in cancer are usually called as onco-microRNA (oncomiR). These miRNAs together with their dys-regulated target genes consist of the oncomiR regulatory networks. Till now, the oncomiRs with known mechanisms are rare and the mathematical models of their regulatory networks are not well established. The results of our research and the current available biological evidences suggest that the miRNA regulation is in multi-scale: some miRNAs has hundreds of target genes, regulating diverse and general biological processes and their target genes are poorly co-expressed; while other miRNAs have much less target genes, regulating some specific biological processes. And also, during the cancer initiation and progression many biological processes with different regulatory scales are dys-regulated. So, we propose that a mathematical of oncomiR regulatory networks should be established based on the principle of multi-scale regulations. Then, the model will be applied on large-scale cancer expression datasets to systematically identify and analyze the oncomiR regulatory networks.