Sorry, I want to ask a naive GO related problem again. If I have the predicted target list of miRs with no P values of difference expression, how could I do functional enrichment analysis with R package GOstats ? Thank you.
GOstats is designed as a package to interact with GO and expression data. I think it may not possible to calculate GO enrichment using GOstats with out p-values from expression studies. You can calculate GO enrichment with out p-value using other packages; for example I have used topGO.
A detailed tutorial is provided in the documentation of topGO. Please refer to Section: 4.2 Predefined list of interesting genes for a detailed example of GO enrichment calculation using predefined list of genes with out a p-value.
You may also want to try the package GeneAnswers. It's pretty easy to use and has some nice visualizations.
I found someone use GOstats to calculate GO enrichment with out p-value:
But using the gene list provided, I do it in the DAVID,but the result are very different(see here and the gene set).
I don't know what make the difference.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
But I use all the target set to do enrichment calculation not a predefinded list(such sample to get in the example to render " allGenes must be a factor with 2 levels"), so could you give me further suggestion?
I am sorry, I can't read Chineese/Mandarin. But I checked the R code. It seems you can implement enrichment calculation with out p-value. I am not understanding your specific question and where you need suggestion. It is not clear to me how can you do an enrichment calculation with out a predefined list. Enrichment is calculating for the list with respect to the background. If list = background, how can you expect enrichment ?
Sorry, Khader, I am a freshman in this area. I think the background should be the all the mouse genes if I predicted the targets as the miRNAs. Do you think so?
No probs Chuangye. Yes. If you are looking at some miRNA gene and would like to calculate the enrichment of these with respect to mouse genome you should use mouse genome as the background. In case if you are looking at microarray experiment, you need to specify the genes in the array platform as your background. Trust this helps.
Sorry, there is a mistake in the last reply of mine "....if I predicted the targets as the miRNAs......" should be "....if I predicted the targets of the miRNAs......" . I think your your reply will be suitable to it either. Thank you very much, Khader.
Good to know that the answer helped. You may accept the answer if you find it useful.