RNA Seq: differential gene expression with multiple factors

[jvelotta]

Hello,

I'm looking to assess differential gene expression using Illumina HiSeq reads. I have a two-factor experimental design (three divergent fish populations each treated at two salinities). I would like to assess differential expression using a two-way model - specifically I want to pinpoint genes for which there are significant interaction effects.

Does anyone know if any of the (many) RNA seq tools available can run a model of this kind (probably a generalized linear model) using these data?

Thanks so much.

[kopi-o]

edgeR and DESeq/DESeq2 can do this and yes, they use GLMs. There are probably other packages too but these are the ones that come to mind.

[jvelotta]

Thanks. Any experience with either of these packages? Are there major differences?

[kopi-o]

I'm a satisfied user of both. I would not say there are major differences between them.

[shi]

Another option is limma. You use its lately developed 'voom' function to normalize your data and then you can analyse your data in the same way as analyzing microarray data.