MOTIVATION: A large amount of data on metabolic pathways is available in databases. The ability to visualise the complex data dynamically would be useful for building more powerful research tools to access the databases. Metabolic pathways are typically modelled as graphs in which nodes represent chemical compounds, and edges represent chemical reactions between compounds. Thus, the problem of visualising pathways can be formulated as a graph layout problem. Currently available visual interfaces to biochemical databases either use static images or cannot cope well with more complex, non-standard pathways.

RESULTS: This paper presents a new algorithm for drawing pathways which uses a combination of circular, hierarchic and force-directed graph layout algorithms to compute positions of the graph elements representing main compounds and reactions. The algorithm is particularly designed for cyclic or partially cyclic pathways or for combinations of complex pathways. It has been tested on five sample pathways with promising results.