With the rapid development of science and technology, large and complex data have been generated in many biological science areas, such as single-cell RNA-Seq, neuroscience and human dynamics. However, the task of analyzing big data itself poses significant challenges. On the one hand, the ultra-large size of datasets renders the application of many statistical methods computationally impossible. On the other hand, with the system being studied getting more complicated, the model setup for some popular off-the-shelf methods may not be applicable anymore. Developing new theoretically justifiable and computationally efficient methods for tackling big data problems from a computational and modeling perspective is the primary motivation for my research. The proposed methods can be widely applied to various scientific disciplines and greatly help scientific development.