This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Differentially expressed gene clustering

Hi all,

I see in many papers that differentially expressed (DE) genes are clustered into several groups. I have no experience with this issue, but as far as I understand, those DE genes with similar expression pattern grouped in several clusters with a given p-value, so we have a general picture of expression pattern, am I right? could you please explain the other benefits of clustering? there are several tools and methods for clustering, how I should select the right method for clustering?

Thank you

differentially expressed gene clustering

If you would like to obtain several clusters of your genes, which show similar expression profile, then you may use ExpressCluster software (http://cbdm.hms.harvard.edu/LabMembersPges/SD.html) in stand-alone version or at GenePattern server. Clustering shows you just groups of genes, which share some similar feature in assumed manner. You may group it by similar expression profile (like ExpressCluster does), neighbouring location on chromosome and so on. The methods for clustering are described in ExpressCluster manual.

2 answers


The clustering you are referring to is for reducing redundancy. For @seta's use, this is the wrong kind of clustering.

"Similar expression pattern" in the context of DE means these genes probably share similar or same function in biology. A special case is driver genes. This kind of analysis will result in more focused results, Here is a tool cogena, co-expressed gene set enrichment analysis I developed. Hope this can be useful.

Thank you for your response. However, I couldn't find documentation for using it, please let me know if it is suitable for a non-model plant?

please find manual and vignette here. Cogena does not support non-human species due to the gene sets, but if you have the plant-related gene set and format it as gmt file, it will work.

Log in to answer this question.