GSA for non-mRNA datasets. (a) The Past (a historical overview of the main GSA methods for non-mRNA datasets): the figure includes the main methods for both genomic range GSA and ncRNA GSA published until 2016. (b) The Future (network approaches to GSA): for genomic data, networks can be used either as links between chromatin regions related to transcription or as linkage disequilibrium clusters, which may redefine the mapping from peaks or SNPs to genes (non-depicted). For ncRNA data, multipartite ncRNA–mRNA correlation networks in tandem with community detection algorithms may become an avenue to understand the different correlation structures between ncRNAs and mRNAs and choose the right gene sets for GSA. Depicted: DE genes (inside a diamond) generate communities on an imaginary network, which may be used as gene sets for integrative GSA of RNA data.