DESEQ2 for ASE

Question

[grant.hovhannisyan]

Hi Biostars,

I have allelic counts of yeast, which I have obtained by mapping RNAseq reads to a diploid genome. Is it legitimate to use DESEQ2 to assess allele-specific expression for this kind of data?

Thanks, Hrant

[Fabio Marroni]

Theoretically it should be legitimate, provided you correctly accounted for all the possible mapping bias you might have when dealing with allele specific expression. However, you can also use some software developed for ASE, such as allim. I don't know if you are expert in ASE, but in case you want to read something you can give a look to this paper, in which an additional method is also presented (many additional paper out there!)

[grant.hovhannisyan]

Thanks for replying. provided you correctly accounted for all the possible mapping bias you might have when dealing with allele specific expression I have mapped the reads using STAR with high stringency (allowing 0 mismatches) to a reference genome with both haplotypes. I guess this removes possible mapping bias problem, right?

[GouthamAtla]

Allowing 0 mismatches would not account for sequencing errors.

[grant.hovhannisyan]

Agree, but don't you think its better to miss a little proportion reads than increase the number of multimaps? But thanks, I will try with 1 or 2 mismatches to see how different the mapping statistics is.

[GouthamAtla]

In anycase, you need to have different biological samples to assess the allele specific expression.

[grant.hovhannisyan]

Do you mean biological replicates? I have 3 for each parent and their hybrid.

Answer 1

[venu]

This might help

Using RNA-seq DE methods to detect allele-specific expression

[GouthamAtla]

moved from comment to answer.