关键词: 玉米, 多组学, 整合网络图谱, 机器学习

Abstract: Life science is entering into the era of big data due to the rapid development of high-throughput omics technology. Multi-omics data such as genome, transcriptome, proteome, metabolome have greatly facilitated dissecting the complex and sophisticated regulatory networks of organisms. Recently, a collaborative team led by Lin Li, Fang Yang and Jianbing Yan from Huazhong Agricultural University constructed the first multi-omics integrative network map of maize. This map comprises over 30 000 genes and 2.8 million network edges at the levels of genome, transcriptome, translatome, and proteome, finally forming 1 412 regulatory modules. Using the integrative network map, the research team successfully predicted and confirmed five new functional genes regulating the development of tiller, lateral organ, and kernel in maize. Based on the integrative map and machine learning, the research team identified 2 651 maize flowering time genes that are enriched in eight candidate subnetworks. The biological functions of 20 flowering candidate genes were further validated using CRISPR/Cas9 gene editing technology and EMS mutants. Furthermore, evolutionary analysis of the integrative network map showed that the two subgenomes of maize had undergone a progressive functional differentiation from the levels of co-expression, co-translation to interactome. The construction of the multi-omics integrative network map represents an important breakthrough in maize functional genomics, which provides a new tool for cloning new genes, identifying novel molecular regulatory pathways, and revealing maize genome evolutionary features. This multi-omics integrative network map is a new key to decode maize functional genomics.

Key words: maize, multi-omics, integrative network map, machine learning