Combined Association Of Expression, Genetic, Clinical Phenotypes

Question

[User 7754]

Hi all,

Can we do an expression-genotype-biomarker/disease association using RNA samples, genotypes, and biomarker/disease outcome from one sample combined with genotypes and biomarker/disease outcome (but no expression data) with another sample? I imagine we can just combine the data for the genotype-expression and also biomarker information, but I am not sure about the disease outcome… Any thoughts? Thank you, -f

Answer 1

[Burlappsack]

Have you thought about using Connectivity Mapping? A quote from the website describes it as such: "This research effort aims to generate a detailed map that links gene patterns associated with disease to corresponding patterns produced by drug candidates and a variety of genetic manipulations"

Answer 2

[Francy]

Hi burlappsack,

thanak you for the link, it seems like an interesting resources. However my question was more a methodological one: would it be possible to merge all of the genetic and phenotype data and do an eQTL association even though we have eQTL, genotype and phenotype data only for a subset of these?

[David Quigley]

@francy: Welcome to the site. This is a question and answer forum, so the correct thing to do if you're responding to someone's question is to use the "add comment" link under their question rather than creating a new answer.

Answer 3

[Larry_Parnell]

You can only perform the association tests using those samples for which you have the appropriate data - that is, the subset you mention. Basically, you are talking about closing a triangle formed by genetic variant, mRNA and biomarker, with the mRNA-biomarker link being the one that adds a lot to the understanding of disease risk/onset/progression.