Reordering based integrative expression profiling for microarray classification
The identification of molecular entities involved in human diseases has been a primary focus of post-genomic biomedicine for pursuing the clinical goals of diagnosis and therapeutic treatment. An emerging perspective in systems biology is that the essential biological roles of molecular entities seem to be well correlated with general molecular network properties. Several types of biological complex networks, including protein interaction networks, have a feature of scale-free networks that relates to fractals (multi-scale self-similarity). Using Alzheimer's Disease (AD) as a case study, we constructed an AD-relevant protein interaction subnetwork. We further developed a computational framework based on Ant Colony Optimisation (ACO) to rank disease network relevant nodes. In this framework, the task of ranking nodes is represented as the problem of finding optimal density distributions of 'ant colonies' on all nodes of the network. Our results also revealed fractal-like properties of the network.