Housekeeping Gene Identifications

Question

[Vitis]

Is there any straightforward way to identify housekeeping genes based on a series of genome-wide gene expression profiles? Literature searches suggest that meta-analysis of large-scale expression data can do the task, but what if the organism doesn't have enough expression data to mine? Is it safe to grab the genes with fold changes bouncing around one (0.9~1.1 range) if comparing all samples pairwise?

[Sean Davis]

I think this answer depends on why you need to know the housekeeping genes. What is the plan for using them once you have them?

[Vitis]

The main purpose is to identify a set of housekeeping genes to help normalize mRNA-Seq expression profiles across tissues and experiments. Much like what TMM normalization does, we'd like to use a set of housekeeping genes to calibrate or accommodate the RNA composition and/or sequencing depth difference across samples.

Answer 1

[Casey Bergman]

Yes, as you suggest using meta-analysis of microarray data can lead to good candidate for housekeeping genes on the basis of expression level stability across time and tissues. You should be able to extend a set of predicted housekeeping genes based on this approach from a species with abundant data to another related species without much microarray data, by finding the orthologs of candidate housekeeping genes. Proof of principle that this approach should work can be found in de Jonge et al (2007) "Evidence Based Selection of Housekeeping Genes", who perform microarray meta-analysis independently in human and mouse and show that:

...novel candidate housekeeping genes identified in the human data set also showed stability in expression in mouse arrays...

HTH, Casey

[Vitis]

Thanks for the suggestions. But identifying orthologs in a genome-wide scale is a great endeavor, I'll think about the best approach before start.

[Casey Bergman]

Depending on your species, finding orthologs may already have been done. Perhaps you should look for other posts on BioStar related to this issue.

[Vitis]

Yeah, I read a lot of discussions on the topic here, and thought I'd take a phylogeny-based approach to really make sure I get the right orthologs.

Answer 2

[Bidossessi Wilfried]

Hi . You can find in the database HT Atlas reliable lists of human and mouse housekeeping genes. http://www.housekeeping.unicamp.br.