Exploring Transcriptional Regulation of Soybean Tissue Development with Machine Learning Method

Abstract

AbstractSoybean is one of the most important crops that is widely demanded by people in daily lives. Measuring the transcriptome of a tissue or condition is a powerful way to detect changes in genetic adaptation. However, it remains difficult to identify the key genes in transcriptional regulation most likely to explain specific traits. Here, we outline a machine learning method that utilizes publicly available soybean RNA-seq data by uncovering conserved expression patterns of genes controlled by transcription factor (TF) / transcription regulator (TR) genes in soybean tissues across time and space under various conditions. In addition to its function in gene expression homeostasis, we can also identify important TF/TR genes related to soybean leaf, stem and root tissue development. Combining with co-expression modules highly expression in the tissue, we also highlight the impact of candidate TF/TR genes in the module in different tissues that may shape the dynamics of soybean development. Together, our results revealed the importance of transcriptional regulatory module analysis in unraveling key traits in the soybean development, in particular those TFs/TRs and their target genes.