Skip to Main Content
Switch-like phenomena within biological systems complicate the inference of gene regulatory networks. In this case, the difficulty comes from the fact that the model cannot be inferred from the mixed unknown contexts directly. It is necessary to identify the dasiapurepsila contexts from the data and given a dasiapurepsila context, subsequently infer a model. In this paper, a wavelet-based approach is addressed for the efficient partitioning of data into different biological contexts. The wavelet transform is a well known tool from the signal processing domain. This approach is able to identify the switches in the various conditions, with much lower computational cost than existing techniques. In order to demonstrate the proposed algorithm, experiments on the basis of simulated sequences and a synthetic sequence derived from real gene networks have been performed.