Biclustering of data matrices in systems biology viaoptimal re-ordering

摘要

In this work we present an optimal method for the biclustering datamatrices in systems biology. Our approach is based on the iterativeoptimal re-ordering of submatrices to generate biclusters. A network flowmodel is presented for performing the row and column permutations for aspecified objective function, which is general enough to accommodatemany different metrics. The proposed biclustering approach is applied toa set of metabolite concentration data and we demonstrate that ourmethods arranges the metabolites in an order which more closely reflectstheir known metabolic functions and has the ability to classify relatedobjects into groups.