I am analyzing RNA-seq data for 5 different bacterial strains, 3 of which are "model" organisms (found in common databases) and 2 of which are not (e.g. not found in KEGG). I've annotated their genomes with BlastKOALA and InterProScan for KEGG and GO terms. 1) Which tools are best to use for functional enrichment analysis where I can provide my own genome annotation? 2) Ideally, for the 3 model organisms, I'd like to compare the output of the analysis for different versions of the annotations (mine vs. in existing databases).
At its most basic, enrichment analysis is just a test for the overlap between two lists based on the hypergeometric distribution. So you could easily implement this yourself. However, a number of available tools also allow you to use custom annotations e.g. Bioconductor/R topGO, the Cytoscape plug-in BiNGO.