Support Vector Clustering of dependencies in microarray data

Structuring genes in groups is possibly useful to gain insight into biological and regulatory processes. We propose an application of Kernel Methods in order to cluster homogeneous features, such as pairs of gene-to-gene interactions. Specifically, we apply Support Vector Clustering (SVC), which is a novelty detection algorithm, to provide groups of similarly interacting pairs of genes in respect to some measure (i.e. kernel function) of their activation/ inhibition relationships. In our approach we take advantage of the adjacency graph obtained from the approximation of a combinatorial optimization problem. The effectiveness of the proposed application is given by comparing the modularity results of the obtained clusters with other standard techniques using a biological data set of microarray experiments