Most protein interactions in organisms can be simplified as networks (PPI), but we can not locate each protein specifically in a single-layer PPI network model which makes it difficult to study the difference between PPI in special regions in an organism and the whole PPI in an organism. We divided the rice cell into several organelles to study the relationship between organelles and cell functions. Moreover, we used different cancer tissues and normal tissues in human body to construct two multilayer PPI and study the underlying mechanics of cancer comparing with the normal tissues.

　　We introduced the mathematical theories of multilayer networks along with the concept of multi-degree, overlap of edge, activity of node, multiplexity, degree correlation, participation coefficient and degree correlation coefficient, which are used to analyze the nontrivial structure and properties in PPIs of rice, cancer and normal tissue.

　　In rice PPIs, we investigated the power-law distribution of edge overlap and node activity, and found important proteins related to functional organelles in cell by analyzing participation coefficient, and organelles of strong correlations or weak correlations by analyzing heat map of layer's degree. In PPIs of cancer and normal tissue, we discovered that there are strong core structure, high homogeneousness and robustness in cancer by multiplexity diagram analysis and alignment of core networks, and the nontrivial distribution of overlap and activity. There are more core proteins in cancer than in normal tissues and various types of cancers have mostly the same core structure along with the degeneration of tissue-specified proteins. It is indicated that the core structure of cancer PPIs is related to cell proliferation, growth and transfer. Moreover, different cancers have their own specified proteins that are probably encoded by different oncogenes which lead to series of gene expressions and biochemical reactions during the formation of the same core network.

　　The former results showed that multilayer networks is a creative and powerful tool in studying complex systems and can be used in biology widely to find nontrivial properties and mechanism of biology phenomenon.