DNA micro-array technology has tremendously accelerated the pace of identification of critical DNA sequence motifs in both prokaryotic and eukaryotic organisms. Sophisticated statistical algorithms, such as the Hidden Markov Model, have been applied to the problem of discovering motifs. Such algorithms have been implemented in a number of bioinformatics software tools including Gibbs Motif Sampler, AlignACE, BioProspector and CONSENSUS. These computer programs have been successfully used to identify co-regulated genes in bacteria and yeast. However, these bioinformatics tools differ markedly in their output formats as well as differing in the semantics of their results. As a result there is no systematic framework that facilitates efficient data exchange, querying, consistency checking and merging of DNA motif profiles, especially when the profiles originate from several sources.

We propose a novel language, called MotifML, that can address these critical problems such that motif profiles from different species identified using different software tools and algorithms can be integrated and compared with each other as well as with data formatted in other bioinformatics languages. The proposed language also addresses the problem of representing interactions among regulatory regions so that they can be simulated and visualized. The adoption of such a language will expedite and formalize the process of uncovering and annotating co-regulated genes at different developmental stages and can bring new insights into our current understanding of biopathways. It will also better facilitate development of tools for analysis, consitency checking, and visualization of DNA motif profiles.