Question

Detection of genes associated with polymorphism

0

Entering edit mode

anna.cot.anna.cot ▴ 30

@annacotannacot-20795

Last seen 8 hours ago

United States

Hi,

I would like to double check whether the the quasi-likelihood method, (glmQLFit and glmQLFTest), are correct to use when looking for genes which expression is affected by a selected polymorphism? I have a SNP of interest and run the glmQLFit which gave me nice logical results. However, the bioinformatician I talked to told me edgeR is not an appropriate tool for this sort of analysis. Is it so? And if the answer is yes- why not?

Thanks!

edger • 703 views

ADD COMMENT • link 4.7 years ago anna.cot.anna.cot ▴ 30

0

Entering edit mode

Thank you Gordon, my data definitely meets the criteria :)

ADD REPLY • link 4.7 years ago anna.cot.anna.cot ▴ 30

score 0 · Answer 1 · 2020-04-21

0

Entering edit mode

Gordon Smyth 52k

@gordon-smyth

Last seen 3 hours ago

WEHI, Melbourne, Australia

edgeR is suitable for most differential read count contexts. If your data generates differential read counts, then edgeR should be appropriate. If it doesn't, then it won't be.

ADD COMMENT • link 4.7 years ago Gordon Smyth 52k

0

Entering edit mode

If we're talking here about eQTLs, I guess the doubt raised by the mentioned bioinformatician was more in computational terms. If you want to find eQTLs between thousands of genes and hundreds of thousands of SNPs, you would have to call edgeR such hundreds of thousands of times, which would be probably computationally very expensive. I wonder whether you have any experience in that context using edgeR or limma. Obviously for one selected polymorphism (SNP or whatever other genetic variant) this is not an issue.

ADD REPLY • link 4.7 years ago Robert Castelo ★ 3.4k