RNA-seq: Expression levels and polymorphisms

[Glongo]

Hello from beautiful Santa Cruz, CA!

I am interested in identifying variable expression levels as well as sequence polymorphism between conspecifics in different environments all in one fell swoop and it seems that RNA-seq is the right tool for this mission. However it seems that detecting variable expression levels requires ample coverage for sensitivity purposes while detecting polymorphisms (i.e. SNPs) doesn't require as much depth. Is it safe to assume that if you obtain sufficient coverage to detect variable expression that you will have the power to detect polymophisms? Are there any models for estimating sufficient sequencing depth to achieve the desired power to detect expression differences? I am working with teleost fishes with ~1Gb genome.

Any insight into the pros and cons of RNA-seq library preps (price and efficacy) would also be greatly appreciated!

Cheers,
Gary

[jparsons]

RNA-seq will miss any noncoding polymorphisms (introns/promoter/etc). You will be also be discarding potential SNPs in lowly-expressed transcripts where your coverage is too low to detect DE. Depending on your exact use case (are you interested in sequence variants which may affect distal sites?) you may be better served by doing both genome sequencing and RNA-seq.

I'm actually having trouble thinking of a situation where you would not want both: polymorphisms within transcripts seem highly unlikely to have anything to do with the expression level of said transcript, unless they affect splice sites.

The other questions are a bit outside my wheelhouse.

[Glongo]

RNA-seq will miss any noncoding polymorphisms (introns/promoter/etc). You will be also be discarding potential SNPs in lowly-expressed transcripts where your coverage is too low to detect DE. Depending on your exact use case (are you interested in sequence variants which may affect distal sites?) you may be better served by doing both genome sequencing and RNA-seq.

I'm actually having trouble thinking of a situation where you would not want both: polymorphisms within transcripts seem highly unlikely to have anything to do with the expression level of said transcript, unless they affect splice sites.

The other questions are a bit outside my wheelhouse.

Hi jparsons,

Thanks for the reply! I should have been more specific, I am not looking for SNPs that correlate to changes in expression but want to be able to detect both variable expression levels and SNPs between individuals from different populations. For instance in a MHC complex gene I'd like to be able to identify SNPs or other polymorphisms between individuals and, at the same time, be able to identity different expression levels of say some metabolic gene.

Cheers,
Gary