gene set enrichment analysis

Question

[nishimalhotra2612]

hi all

i have one doubt regarding the gene set enrichment analysis method i have a list of upregulated genes and i have not done the differential gene expression analysis but instead i did rank prod analysis from which i obtained the list of upregulated genes so now when i m searching about the ways to analyse them i m not understanding as most of them are analysing the result from deseq2 and i m looking for the way to analyse my gene list so is there any tutorial or any material which works only on gene list

thanks

Answer 1

[ATpoint]

So basically you have a list of genes and now want to know whether they're enriched for some terms or processed? I like g:profiler2 for these kinds of analysis. It has an R package if you want to automate the task.

[nishimalhotra2612]

thanks for suggesting i tried to use that but the thing is that i have my gene list in form of csv the name is "upregulated.csv"and it has 720 genes so when i try to input my files and run the code it says no result to show. i tried to found out the format but they only mentioned the types of input not the format

so in what format i can input the data if you have any idea

thanks

[ATpoint]

Assuming the formatting of the genes was correct (please show head upregulated.csv) it can simply mean that there are no significant enrichments.

[nishimalhotra2612]

[nishimalhotra2612]

this is how the head(upregulated.csv) looks like

[ATpoint]

Yeah, if you indeed feed these IDs into the tool and it returns nothing then there is simply no statistical enrichment. Try relaxing the FDR threshold, maybe that helps. You can do that with the R package.

[nishimalhotra2612]

Okkk thanks I will try that as well and see which fits better

Answer 2

[Marco Pannone]

In your case, I would suggest performing an over-representation analysis (ORA) rather than GSEA. One quick way to perform ORA is on WebGestalt.

If you want to perform a proper GSEA, you need to have a full count matrix with normalized read counts per each gene as rows and all samples as columns. Then you can run GSEA, as described here (https://www.gsea-msigdb.org/gsea/index.jsp).

[nishimalhotra2612]

Okk thanks for suggesting I will try this

Answer 3

[Jeremy]

In addition to ATpoint's suggestion of g:Profiler, you can try PANTHER, agriGO, and DAVID. Each of these tools takes a list of genes and returns gene set enrichment analysis results.

PANTHER

agriGO

DAVID

[nishimalhotra2612]

Thanks i will take a look

Answer 4

[nishimalhotra2612]

hello all i tried to use gprofiler2 and i got the result i m sharing my way on what i did to input multiple queries so that it can be helpful to others

so as it is being said in the gprofiler tutorial that we have to create a named list of vectors so i did that its very basic code but it can be helpful to the people who are new to bioinformatics

first input the csv file df<-read.csv("upregulated genes.CSV")

select the column consisting genes or gene ids df1<-df$x

converting the selected column to vector vectordata<-as.vector(df1)

then creating the list of that vectors gene_list<-list(vectordata)

and now input the gene_list into gprofiler

hope it helps!