EAT-Rice: A predictive model for flanking gene expression of T-DNA insertion activation-tagged rice mutants by machine learning approaches

摘要

Author summary Among all the food crops, the rice is one of the staple foods in the human population, especially in Asia. However, the human population increases rapidly and the cultivated areas decrease in these decades. To solve the food crisis in the future, the rice researchers devote themselves to study on the gene function to increase the rice yield and stress tolerant ability. There are around 39000 annotated genes in rice, so scientists are hard to survey the gene functional because of the complexity and interactivity among the genes. Therefore, scientists put into a lot of manpower and funds into the field. The T-DNA (Transfer DNA) activation-tagging biotechnology has been wildly used on studies of rice gene function, however, it might influence the flanking genes expression when T-DNA inserted into the rice genome randomly. Thus, it will take lot of time for the researchers to validate the activation of genes by T-DNA enhancer. In these decades, as the increase of the biological data accumulation, the extraction of hidden information from this data is getting more and more important. To assist rice biologists in rapidly focusing the target gene affected by T-DNA. The application of machine learning methods in artificial intelligence (AI) and the establishment of prediction tool with biological data construction to correctly identify and classify target genes are of great significance in both theory and practice.