Metabolic networks have been studied for decades, and sophisticated computational frameworks are needed to augment experimental approaches to harness these complex networks. BNICE (Biochemical Network Integrated Computational Explorer), a computational approach for the discovery of novel biochemical pathways, overcomes many of the current limitations. BNICE and similar frameworks can be used in a myriad of different areas: (i) design of novel pathways for metabolic engineering and bioremediation; (ii) retrosynthesis of metabolic compounds; (iii) evolution analysis between metabolic pathways of different organisms; (iv) analysis of metabolic pathways; (v) mining of 'omics' data; and (vi) selection of targets for enzyme engineering. We will discuss the issues and challenges in building such a framework and the gamut of applications that they can offer.