Development of more advanced statistical and computational methodologies is required to keep pace with new experimental technologies in molecular biology and genetics. Challenges to be met include dealing with increased volumes of data, improving the robustness of inference to noise in the data, and enhancing the models to make them biologically more realistic.