Algoritmos alternos de bajo coste para la comparación de rutas metabólicas en plantas

Abstract

Metabolic pathways provide key information to achieve a better understanding of life and all its processes; this is useful information for the improvement of medicine, agronomy, pharmacy and other similar areas. The main analysis tool used to study these pathways is based on the idea of pathway comparison, using graph data structures. Graph comparison has been defined as a computationally complex task. We propose two algorithms with different approaches which simplify the problem of comparing pathways represented as graphs. The first algorithm consists in the transformation of a two-dimensional graph structure to a one-dimensional structure, and thus aligning the corresponding data using a reduced 1D structure. The second algorithm consists in performing a pair analysis between graphs, that is to say a relation of 2 equal nodes present in both graphs, and thus eliminating all similarities, finally, showing these differences to the user. Our results show evidence of a quick, simple and effective way to resolve the described problem. The mechanism proposed in algorithm 1 can be used as a prior evaluator to predict good comparisons in case a deeper analysis is desired. We show that the loss of information or precision does not affect much the result, which is to give the user a similarity score between the two analyzed pathways. For algorithm 2 the proposal is to offer the expert an additional point of view for his evaluation of the pathway in question. In this case, no score is provided but the listed differences.