Hi, First of all, the RPKM count normalization is not recommended for Differential gene expression analysis as RPKM values are not comparable between samples. For such analysis, it is recommended to work with raw read counts then normalize the data to DESeq2’s median of ratios or trimmed mean of M values (TMM).

Learn more about count normalization at: https://hbctraining.github.io/DGE_workshop/lessons/02_DGE_count_normalization.html#2-create-deseq2-object

If you use DESEQ2, the input data just needs a raw count table, and the DESEQ2 workflow will do the rest.

Deseq2 package http://www.bioconductor.org/packages/release/bioc/vignettes/DESeq2/inst/doc/DESeq2.html

Tutorial: https://www.reneshbedre.com/blog/deseq2.html

https://lashlock.github.io/compbio/R_presentation.html

Usually, RNA differential expression analysis is followed by Gene Set Enrichment Analysis(GSEA). Most GSEA analyses will have gene annotation. A quick google search will give you many different packages such as Enrichr, clusterProfiler, Gage/pathview etc.

or you can also stop after differential analysis and not proceed further for GSEA (depends on your aim as some pathways are not well characterized). In this case, you can manually annotate your orfs/genes/transcripts normal gene annotation process.

Other widely used differential analysis tools are limma, EdgeR and others.

Shailesh Nair HI thank you for the reply, I believe the RPKM values are the result of differential gene expression analysis (already done then handed off to me). Which in this case I do not need to use DEseq2, edgeR etc. The issue I am having is what would be the best way to utilize/modify the data in the columns so that I can easily perform further functional analysis in R such as GO, GSEA etc as I understand as the input format I have is uncommon in the vignettes/tutorials I have seen (is it possible to convert this to log2foldchange?). Thank you!

Hi,

The three GSEA tools I use the most often are:

DAVID https://david-d.ncifcrf.gov/tools.jsp (submit your list of Ensembl IDs in the "Upload" section, and select "ENSEMBL_GENE_ID" from the drop-down list, then follow the instructions and choose the Gene Sets to use).

MSIGDB http://www.gsea-msigdb.org/gsea/msigdb/ (once registered, submit your list of Ensembl IDs in the "Investigate" section, and select the Gene Sets you want to use).

ENRICHR https://maayanlab.cloud/Enrichr (submit your list of Gene Names)

Choosing too many unnecessary Gene Sets to test against will weaken your p-values as it increases the multiple-testing burden. Also, categories such as Transcription Factors and Microarray targets may swamp your results with pages of hits you can't interpret. (or maybe these are exactly what you're looking for!). It's best to start with just the GO ontologies and KEGG pathways and repeat with more sets if the initial results aren't helpful.

Also, bear in mind the "background universe". For example, a set of genes differentially regulated in brain tissues will almost always give "Neurogenesis" as a prominent hit - this does not mean that neurogenesis was a differentially regulated pathway in your experiment, only that the pathway is enriched in brain-expressed genes. DAVID allows you to submit a background list (in this example it would be all brain-expressed genes rather than the entire genome) to counter this. You probably don't have this option if you were only given the differentially expressed genes, but it's worth being aware of.

Mikan Minatozaki Hi, I am not sure which DE analysis gave you RPKM values, but the current DE analysis does not recommend RPKM normalization for comparisons across samples which will give you biased results.

Even still if you want to proceed, you can try using :

RPKM function in edgeR

Limma https://bioconductor.org/packages/release/workflows/vignettes/RNAseq123/inst/doc/limmaWorkflow.html

MSIGDB https://www.gsea-msigdb.org/gsea/index.jsp

I would highly recommend that you re-do your DE analysis with appropriate normalization such as TMM or median-of-ratios method

Try this tools

David : https://david.ncifcrf.gov/

Webgestalt: http://www.webgestalt.org/

Funrich: http://www.funrich.org/

try this tool https://biit.cs.ut.ee/gprofiler/gost