Sorting bedgraph using gnu sort instead of bedtools sort

[johapark]

Related to this issue #344, I frequently encounter the out-of-memory error during sorting bedgraph files when the file size is large. Can we use GNU sort instead of bedtools sort?

I noticed that you previously used GNU sort but later decided to switch to bedtools sort (#290). However, this issue can be simply solved by adding the following line export LC_COLLATE=C and the job can be done more quickly with less memory.

Is there a particular reason for switching to bedtools sort? If not, could we use GNU sort instead for quicker sort with less memory?