Differential gene expression using R I am working on RNA Seq data analysis to get differential gene expression between 2 conditions. I am using ballgown package on R, and successfully loaded the data into R.
However, I do have these queries after my progress:

*

*Is it necessary to remove low variance transcripts while doing differential gene expression? And why?

*Why do we need to remove low gene abundance & low variance transcripts?

*How do I get gene name and gene id without stattest() function on R using ballgown?

Thanks in advance!
 A: I think bioconductor will be a good start to get a handle on this.


*excluding genes with poor count/abundance is suggested as one never know if they are an artifact or in real.
You can start with deseq paper to understand more on variance's technical aspect: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2
http://bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html
I cannot answer for Q1 and Q2.
A: With respect to Q1, the problem of multiple comparisons looms over this type of study, so there's an advantage to cutting down on the number of genes that you are formally evaluating in the analysis. If a transcript's expression shows little variance among samples it is unlikely to provide much information in a differential-expression study. If there's little variance among samples there's unlikely to be much differential expression between conditions. If you included all transcripts you would have to be more stringent in the multiple-comparisons correction and thus be more likely to miss true positive results. In general, when there are a lot of potential predictors in a model or many outcomes that are being measured, removing low-variance characteristics is a useful and principled way to focus attention on the characteristics that are most likely to matter.
The answer from Death Metal handles Q2 pretty well (+1).
Q3 is about non-statistical details of a particular software function and thus is off-topic on this site.
