Anther development in Arabidopsis, a popular model plant for plant biology and genetics, is controlled by a complex gene network. Despite the extensive use of this genus for genetic research, little is known about its regulatory network. In this paper, the direct transcriptional regulatory relationships between genes expressed in Arabidopsis anther development were predicted with an integrated bioinformatic method that combines mining of microarray data with promoter analysis. A total of 7710 transcription factor-gene pairs were obtained. The 80 direct regulatory relationships demonstrating the highest confidence were screened from the initial 7710 pairs; three of the 80 were validated by previous experiments. The results indicate that our predicted results were reliable. The regulatory relationships revealed by this research and described in this paper may facilitate further investigation of the molecular mechanisms of anther development. The bioinformatic method used in this work can also be applied to the prediction of gene regulatory relationships in other organisms.
