Question: KEGG pathway analysis on a predefined set of genes (with my own annotation)?
0
2.9 years ago by
miyakokodama0 wrote:

I have a predefined set of interesting genes, along with their corresponding GO IDs and KEGG terms. According to the tutorial, I used TopGO and obtained significantly enriched GO terms using the Fisher's exact tests.

I am hoping to do something similar to perform KEGG pathway analysis, however I haven't been able to figure out how to do that. Does kegga() function in the limma package deal with a list of predefined genes with my own annotation?

Miyako

Note: KEGG mapper online to find genes involved in particular pathways, but am hoping to get a p-value for each KEGG term.

limma kegga • 769 views
modified 2.9 years ago by Gordon Smyth38k • written 2.9 years ago by miyakokodama0
Answer: KEGG pathway analysis on a predefined set of genes (with my own annotation)?
1
2.9 years ago by
Aaron Lun24k
Cambridge, United Kingdom
Aaron Lun24k wrote:

Examining ?kegga may reveal some clues as to the expected input:

      de: a character vector of Entrez Gene IDs, or a list of such
vectors, or an ‘MArrayLM’ fit object.
Answer: KEGG pathway analysis on a predefined set of genes (with my own annotation)?
0
2.9 years ago by
Gordon Smyth38k
Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia
Gordon Smyth38k wrote:

It's pretty straightforward. For example:

> GeneID
[1] "8985" "7423" "8200" "5439" "3710" "1364" "1386" "8323" "3952" "5879"
> library(limma)
> ke <- kegga(GeneID)
> topKEGG(ke)
Pathway   N DE    P.DE
path:hsa05205                              Proteoglycans in cancer 205  3 0.00242
path:hsa04918                            Thyroid hormone synthesis  74  2 0.00453
path:hsa04060               Cytokine-cytokine receptor interaction 265  3 0.00502
path:hsa04925                  Aldosterone synthesis and secretion  82  2 0.00554
path:hsa04911                                    Insulin secretion  85  2 0.00594
path:hsa04972                                 Pancreatic secretion  96  2 0.00753
path:hsa04915                           Estrogen signaling pathway 100  2 0.00814
path:hsa04933 AGE-RAGE signaling pathway in diabetic complications 101  2 0.00830
path:hsa04922                           Glucagon signaling pathway 103  2 0.00862
path:hsa04151                           PI3K-Akt signaling pathway 341  3 0.01012
path:hsa04670                 Leukocyte transendothelial migration 118  2 0.01121
path:hsa04728                                 Dopaminergic synapse 130  2 0.01349
path:hsa05200                                   Pathways in cancer 397  3 0.01533
path:hsa04310                                Wnt signaling pathway 143  2 0.01618
path:hsa04932            Non-alcoholic fatty liver disease (NAFLD) 151  2 0.01793
path:hsa04390                              Hippo signaling pathway 154  2 0.01861
path:hsa04022                           cGMP-PKG signaling pathway 168  2 0.02193
path:hsa04510                                       Focal adhesion 203  2 0.03123
path:hsa05169                         Epstein-Barr virus infection 204  2 0.03151
path:hsa05203                                 Viral carcinogenesis 205  2 0.03180

I do wonder however what you mean by "with my own annotation"? There is no way to use your own association of KEGG terms with genes unless you do this for all genes, not just for your set of interesting genes.

Hi Gordon,

Thanks so much for your reply. I think my wording was unclear - I work on a non-model plant species, and I have an input, which is something like this:

GeneID KEGG_ID
 108225144
 K00121
 108197875
 K00297
 108224641
 K00309
 108220854
 K00328
 108223815
 K00422
 108214543
 K00430
 108217705
 K00430
 108204546
 K00430

where I have a list of deferentially expressed genes along its corresponding KEGG term (some genes have the same K number). Using kegga(), I'd like to perform KEGG enrichment analysis. Do I use gene.pathway and pathway.names arguments to run kegga()?

Please let me know if anything is unclear, and many thanks in advance.

Miyako