I should really know this by now, but I can't recall how to cleanly subset a SummarizedExperiment/GRangesList by seqnames. 

Here I gave some bad example code just now, though I know there is something better:

C: remove X and Y chromosome genes in RNA-seq data using DESeq2 pipeline

I searched here first and found:

A: GRangesList filtering

But using Valerie's code there gives:

> grl[seqnames(grl) == "chr2"]

GRangesList object of length 3:
$gr1
GRanges object with 1 range and 0 metadata columns:
      seqnames    ranges strand
         <Rle> <IRanges>  <Rle>
  [1]     chr2    [3, 6]      *

$gr2
GRanges object with 0 ranges and 0 metadata columns:
     seqnames ranges strand

$gr3
GRanges object with 1 range and 0 metadata columns:
      seqnames ranges strand
  [1]     chr2 [4, 9]      *

Already, a better solution than my as.character() hack is to throw a elementLengths on top:

> elementLengths(grl[seqnames(grl) == "chr2"]) > 0
  gr1   gr2   gr3
 TRUE FALSE  TRUE

But there must be a better way I'm forgetting.

> sessionInfo()
R version 3.2.0 Patched (2015-04-26 r68264)
Platform: x86_64-apple-darwin10.8.0 (64-bit)
Running under: OS X 10.8.5 (Mountain Lion)

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats4    parallel  stats     graphics  grDevices datasets  utils     methods   base

other attached packages:
[1] GenomicRanges_1.20.3 GenomeInfoDb_1.4.0   IRanges_2.2.1        S4Vectors_0.6.0
[5] BiocGenerics_0.14.0  testthat_0.9.1       devtools_1.7.0       knitr_1.10
[9] BiocInstaller_1.18.1

loaded via a namespace (and not attached):
[1] XVector_0.8.0 tools_3.2.0

Martin already answered this here:

C: remove X and Y chromosome genes in RNA-seq data using DESeq2 pipeline

Thanks very much Michael and Martin.  I used the codes below to subset the dds object and it worked without any problem.

seqnames(dds)
dds.sub <- dds[ all(! seqnames(dds) %in%  c("chrX", "chrY"))]