PMC:1481596 / 5816-7875 JSONTXT

{"target":"https://pubannotation.org/docs/sourcedb/PMC/sourceid/1481596","sourcedb":"PMC","sourceid":"1481596","source_url":"https://www.ncbi.nlm.nih.gov/pmc/1481596","text":"In this work, we model a functional module as a union of overlapping dense subnetworks called here functional groups. A functional group is either a maximal clique (typically representing a protein complex) or a set of alternative variants of such complexes/cliques. As components of a larger functional module, functional groups are not assumed to be well separated and can have significant overlaps. Intuitively, if a functional module performs a function that requires a sequence of steps (like in the case of a signaling pathway) then we would like functional groups to be snapshots of protein associations at these steps. We propose a new method for identifying and representing overlapping functional groups in a functional module. Furthermore, if the module corresponds to a dynamic process that requires certain complexes (or more generally functional groups) come into contact in a specific order, our method attempts to discover this order. Our method is motivated by a fundamental result for chordal graphs [17], which states that every chordal graph has the so called clique tree representation. However, not every protein interaction network is chordal and not every functional group is a clique. Therefore, we developed a graph-theoretical framework that enables automatic construction of a tree-like representation, analogous to the clique tree representation, for much broader family of graphs. We call this representation the Tree of Complexes representation. The nodes in the tree are functional groups, and for every protein, the set of functional groups that contain this protein forms a single subtree. The \"single subtree\" requirement restricts significantly the way in which the nodes of the tree can be interconnected. As a consequence, this representation shows a smooth transition between functional groups and allows for tracking a protein's path through a cascade of functional groups. Therefore, depending on the nature of the network, the representation may be capable of elucidating temporal relations between functional groups.","tracks":[]}