If you want to see which KEGG or REACTOME pathways are enriched in your list of genes, a simple approach is to use goseq. In the manual they explain how to do the enrichment analysis with custom (other than GO) databases such as KEGG or REACTOME. In my opinion, these results are always less informative than GO enrichment. The number of pathways are limited (compared to the number of GO terms), and GO terms may lead to novel insights (with pathways usually not).

You can also upload your gene list to MSigDB and look for Pathway enrichment.

With GeneSCF batch run, you can try all those database with single run. Selecting the database depends on your experiment or study or interest.

KEGG always gives you broad perspective of pathway terms (metabolism, gene regulation, cellular processes, disease and signaling pathways).

In Gene Ontology, you can have individual terms in-detail like sub processes of these broad categories such as Biological Process (GO_BP), Molecular Function (GO_MF) and Cellular Component (GO_CC).

Reactome has database with sub-pathways of broad terms (broad terms as in KEGG). Reactome has advantage by covering more number of proteins.

More details in GeneSCF article.