An Agent-Based Approach to Mine Transcriptional Regulatory Elements

The study of transcriptional regulation mechanisms is one of current research issues in post-genomic era. However, the number of the known regulatory elements is rarely limited, and the accuracy of the state-of-the-art identification methods is still far from satisfactory. Therefore effective information, personalized service are critical features to be provided in a system, but the existing systems ignore overall performance issues. Agent technology has been used in bioinformatics. But little consideration has been taken into regulatory elements mining. In this paper, the first multi-agent-based system TReMAgent for mining regulatory elements is presented, in which novel algorithms are developed to acquire superior accuracy and integrate with existing tools utilizing agent technology to realize the collaboration and flexibility. In TReMAgent, the workflow plans will be provided automatically to meet with personalized service and databases are integrated to satisfy dynamic mining. The experimental results tested on real data sets show that TReMAgent can collect more quantities and superior accuracy elements and exhibit effective, personalized and flexible services.