Hi everyone:

I ran into issue when I tried to split relatively big GRanges objects by given threshold value, and my approach very slow to reach expected result. My approach could work fast if I used data.frame, but I trust the GRanges object could work well dealing with genomic interval. Can anyone give me possible suggestion to speed up splitting GRanges objects relatively fast? How can make this happen? Any idea ?

> length(gr)
[1] 36678

I tried this way :

lapply(gr, function(x) split(x, c("keep", "saved")[(x$p.value <= 1e-08)+1]))

but doing this way is so slow and output format is undesired, so this motivate me to find out other approach. Which way I can facilitate above process? what if GRanges objects unexpectedly large, and still need to split up by comparing its metadata with given threshold, what should I do ? Any help is appreciated. 

my expected output could be list (the skeleton of the output):

gr
 gr$keep
 
 gr$saved

Here is the session info:

> sessionInfo()
R version 3.3.1 (2016-06-21)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)

locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252   
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

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

other attached packages:
[1] metap_0.7             rtracklayer_1.33.12   GenomicRanges_1.25.94 GenomeInfoDb_1.9.14  
[5] IRanges_2.7.17        S4Vectors_0.11.18     BiocGenerics_0.19.2  

loaded via a namespace (and not attached):
 [1] lattice_0.20-33             XML_3.98-1.4                Rsamtools_1.25.2           
 [4] Biostrings_2.41.4           bitops_1.0-6                GenomicAlignments_1.9.6    
 [7] grid_3.3.1                  zlibbioc_1.19.0             XVector_0.13.7             
[10] Matrix_1.2-6                BiocParallel_1.7.9          tools_3.3.1                
[13] Biobase_2.33.4              RCurl_1.95-4.8              SummarizedExperiment_1.3.82

Why not just do:

split(gr, ifelse(gr$p.value <= 1e-08, "saved", "keep"))