Future developments of g:Profiler will focus on providing biologically more relevant results to the users. On one hand it needs analysis and adaptation of alternative gene set enrichment methods. Since the functional terms are often not independent from each other, we need methods that are capable of taking into account dependencies between the terms and the overall graph structure of the ontologies. Some enrichment tools, such as the topGO package in R, propose methods that take the graph topology of GO into account while testing the enrichment (39). However, this is a remaining challenge that needs further research. On the other hand, the personalized medicine initiatives generate enormous data sets about SNPs and enrichment analysis is moving from gene level to single nucleotide level when studying associations with traits. Therefore, we aim to expand our g:SNPense tool to make use of the large datasets from GWAS and eQTL analyses and to put their results into context and highlight the statistically significant relationships from there.
