Title: Inferring networks from next generation sequencing data
Authors: Thi Kim Hue Nguyen - University of Padova (Italy) [presenting]
Monica Chiogna - University of Bologna (Italy)
Abstract: MicroRNAs (miRNAs) have been reported to play a pivotal role in regulating key biological processes, for example, post-transcriptional modifications and translation processes. Some studies revealed that some disease-related miRNAs can indirectly regulate the function of other miRNAs associated with the same phenotype. Hence, studying the interaction pattern of miRNAs in some conditions might help understand complex phenotype conditions. Inferring the interaction pattern is a challenging task, as data measuring miRNA expression are usually high dimensional, discrete, possibly showing a large number of zeros and measured on a small number of units. From a technical point of view, the interactions among miRNA are well represented by a graph, where miRNAs and their connections are, respectively, nodes and edges. We propose a new algorithm for learning the structure of undirected graphs for count data, called PC-LPGM, and we prove its theoretical consistence in the limit of infinite observations. The proposed algorithm shows promising results when compared to some competitors using simulated data. Moreover, it provides biologically interpretable results when applied to real data downloaded from The Cancer Genome Atlas portal.