Principles, methods and applications of gene regulatory network inference

The gene regulatory network is a complex biological network that controls gene expression in cells. It is also a frontier and a hot topic in systems biology research. It has important biological significance in revealing a series of biological problems, such as growth and development, aging, immune response, and disease occurrence. In this context, this paper reviews the relevant methods and techniques for constructing GRN using statistics, traditional machine learning, and deep learning, and discusses their advantages, disadvantages, and applicable scenarios. Taking the model plant Arabidopsis as a case study, the basic research framework of GRN inference is demonstrated, that is, representative algorithms in statistics, traditional machine learning, and deep learning are selected to construct the GRN, and their effects on predicting GRN are evaluated, and the optimal algorithm for constructing Arabidopsis GRN is screened. This paper aims to provide a reference for GRN inference methods and application research.