Graphical layout techniques serve a vital part in systems biology to enhance understanding and visualization of chemical reaction pathways in our body. Metabolic networks have particularly complex binding structures, making its graphical representation challenging to comprehend. For the purpose of legibility, reducing graph complexity in metabolic networks is crucial when working with large number of nodes and edges. This paper introduces a node abstraction algorithm that treats metabolic pathways as hierarchical networks and considers reactions between compound pairs - the equivalent of node pairs in the context of biological networks - as an elastic parameter for reaction compression in an automated way. Substrates and products that locally compose reactions with low connectivity were reduced, and cyclical or hierarchical pathways were aligned according to their structural composition.