Hi,
I'm a newbie to bioinformatics so apologies for the naivety of my question. I'm doing RNASeq analysis on a list of genes (bacterial genome) under one condition vs another. After using the DESeq2 package in R, I have a list of up+downregulated DE genes. I did gene enrichment analysis on them using PANTHER GO and got back a list of significant pathways (p<0.05 with no correction method).
Next, I've inputted the genes under each pathway that PANTHER outputted into KEGG search&color mapper and have now have a set of several pathway maps, colored based on the genes I inputted. Just to clarify I've a set of maps that color the genes that are downregulated and another set of maps that color the genes that are upregulated.(Some of the these maps cross over i.e I have one map with 1 gene upregulated under the TCA cycle and another with 3 genes downregulated under the TCA cycle). The pathways fall under metabolic pathways, amino acid biosynthesis and nucleotide biosynthesis etc.
I'm a bit stuck on where to go from here. Ideally I'd like to have the significant pathways belonging to say amino acid biosynthesis all together on the one map with upregulated genes colored in one color and downregulated genes in another. Any suggestions on websites/apps to use? Or is there a method in KEGG to do this? Thanks in advanced for your help!
Are you able to put a hypothesis together based on the genes you have and the pathways they seem to belong/map to? Are the results tracking along with what you were expecting to see based on your experimental design?
Yes protein families, biological processes and pathways are what I would expect given my experimental conditions. I want to represent this data clearly now. Is pathview in R a good package for this?
It looks good if you want to actually visually see where your genes are within the pathway. Both clusterProfiler and enrichR are both quite good for just making visualizations of your pathway enrichments.
What format are the geneIDS in geneList of the package DOSE that they use as an example in clusterProfiler? e.g: head(names(geneList)) [1] "4312" "8318" "10874" "55143" "55388" "991" . Also if I'm using clusterProfiler for visualisations is that redoing the GO analysis I've already done in PANTHER? I can't use enrichR because it doesn't support my organism. Thanks very much for your help on this.
They are Entrez IDs. You can convert from many other formats with the package's
bitrfunction. If you run it in full, yes, it would be doing enrichment analysis on its own (though it should be pretty close to what you already got from PANTHER depending on how up to date PANTHER's annotations are).I'd recommend reading the package's documentation to get a better idea of what it can do.
In addition, GSEA is very useful and popular for this sort of analysis.