NES2RA
Gene network expansion is a task of the foremost importance in computational biology. Gene network expansion aims at finding new genes to expand a given known gene network. The expansion lists were computed applying the NES2RA algorithm (Asnicar et al., 2016) to a pre-processed transcriptomic dataset. NES2RA is based on the PC-algorithm, named after its authors Peter and Clark (Spirtes and Glymour, 1991), a gaussian graphical model (GGM) that finds causal relationships from observational data. The PC-algorithm is based on a systematic test for conditional independence to retain significant relations between pairs of genes. It starts from a fully connected network and removes interactions between genes, if it finds a set of genes that supports that interaction (i.e., separation set). From a mathematical point of view, the test for statistical independence between the genes a, b conditioned by a set of genes S is driven by the estimation of the partial correlation ρ (a,b∖S). As the exhaustive exploration of all the subsets of conditioning genes is computational impossible, the PC-algorithm takes into consideration only a limited number of those sets, as described in Algorithm 3 in Asnicar et al. (2016). At the same time, the NES2RA algorithm, developed to cope with this computational complexity, randomly divides the dataset into tiles of equal number of genes (subsetting), where each tile always includes all the genes of the LGN, to be then processed by the PC-algorithm. The random subsettings of all the genes in the genome are repeated for a given numbers of iterations. The NES2RA algorithm currently runs as part of the gene@home project which relies on thousands of volunteers’ computers by means of the BOINC system (Anderson, 2004) within the TN-Grid platform (Asnicar et al., 2015). This approach has been successfully applied to Vitis vinifera transcriptomic data taken from Vespucci database to expand LGN related to climate changes (Malacarne et al., 2018).
OneGenE
OneGenE aims to expand each single gene in an organism. OneGenE systematically runs single-gene NES2RA expansions with fixed parameters, and then combines them afterwards to simulate LGN expansions. In this way the single-gene expansions can be used multiple times, thus effectively reducing the computational effort (Asnicar et al., 2019,Blanzieri et al., 2020). As a consequence the results are obtained in real time permitting the user to interact with the system in order to define the LGN of interest.