Can somehow GSEA tell me which gene set is most significantly enriched in our set of analysis. It gives ES, FDRp values. Alternatively, Is there any tool that will give me a ranking of different pathways enriched in my list of genes.
Can somehow GSEA tell me which gene set is most significantly enriched in our set of analysis. It gives ES, FDRp values. Alternatively, Is there any tool that will give me a ranking of different pathways enriched in my list of genes.
Enrichment ranking doesn't mean one pathway is more affected than other. There is no way, in my opinion, to quantify change in pathway just from GSEA. Enrichment score gives you a hypothesis that some pathways are affected based on genes in that pathway either up or down-regulated.
For example, if genes A, B and C are involved in a pathway X; if they are affected then GSEA will tell you that pathway X is affected. If A and B activates X and C inhibits; that directional regulation will not matter in GSEA suggesting it to be an affected pathway or direction of it being affected. If all A, B and C are positively regulated GSEA will suggest geneset for pathway X is upregulated and vice versa.
The significance of a pathway being affected is based on statistical test that basically finds if geneset involved in regulating pathway X are significantly upregulated or downregulated when compared to a random chance.
You still have to find the direction of regulation by actually looking at what are the important regulators of pathway X are significantly affected and how do they regulate the pathway itself.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Moved to answer.
The paper @ https://www.tandfonline.com/doi/full/10.1080/21623945.2019.1649578 has concluded based on just gene ratio what is most significant. Is it acceptable?
It does not explain what is that gene ratio. In one figure it mentions gene ration and in second it mentions, genes affected by genes in the pathway geneset. I am sure you get a better enrichment values from GSEA analysis. I can only say that published resources are likely to be accepted more than in house, unless you specify the statistics behind it. In the current example, I find it vague. Why did they use gene ratio and not enrichment or pvalue. Rest is upto preferences.