Samtools sort vs Picard Sortsam

[rahul]

Hi all,

Which is a better pick for sorting large SAM files in terms of memory requirement and run time: sortSam from Picard or Samtools sort function. Could you please direct me to the thread if it is already discussed and i missed it.

Thank you,

Rahul

[ymc]

I want to know SortSam vs samtools sort, too.

I think MarkDuplicates is better than rmdup because the former can handle duplicates across break points.

[sklages]

Why don't you just try?

[ymc]

Why don't you just try?

Because they give the same results after I try. I think there might be differences in edge cases, for example, when it is piped from bwa-mem. But I don't know what the edge cases are.

[sklages]

So if you get the same results in terms of mem and runtime for huge files, just decide for the one you're more familiar with. Edge cases are usually very specific, so if you don't encounter any problems with your regular files, no need to worry :-)

[swbarnes2]

I think Picard's sortSam puts a line in the header indicating that the file is sorted, which samtools does not. But for lots of applications this doesn't matter.

[dpryan]

I think Picard's sortSam puts a line in the header indicating that the file is sorted, which samtools does not. But for lots of applications this doesn't matter.

Are you sure about that? The header is grabbed from the first tmp file written.