We consider the problem of aligning two metabolic pathways. Unlike traditional approaches, we do not restrict the alignment to one-to-one mappings between the molecules of the input pathways. We follow the observation that in nature different organisms can perform the same or similar functions through different sets of reactions and molecules. The number and the topology of the molecules in these alternative sets often vary from one organism to another. In other words, given two metabolic pathways of arbitrary topology, we would like to find a mapping that maximizes the similarity between the molecule subsets of query pathways of size at most a given integer k. We transform this problem into an eigenvalue problem. The solution to this eigenvalue problem produces alternative mappings in the form of a weighted bipartite graph. We then convert this graph to a vertex weighted graph. The maximum weight independent subset of this new graph is the alignment that maximizes the alignment score while ensuring consistency. We call our algorithm SubMAP (Subnetwork Mappings in Alignment of Pathways). We evaluate its accuracy and performance on real datasets. Our experiments demonstrate that SubMAP can identify biologically relevant mappings that are missed by traditional alignment methods and it is scalable for real size metabolic pathways.