Sort BAM files by reference, and then by read name within each reference?

Question

[vitor.rca]

Is there an efficient way to sort a BAM file by chromosome, and then by read names within each chromosome?

I could split the BAM by chromosome, sort each resulting BAM by read names, and finally merge the list of BAMs. But I wonder if there is a single and straightforward command to do that.

[lieven.sterck]

I'm not even sure you can sort BAM on reference ? or do you mean on position ?

[vitor.rca]

The default behavior of samtools sort does that: "When the -n option is not present, reads are sorted by reference (according to the order of the @SQ header records), then by position in the reference". I am asking how to sort by reference, then by read name.

[lieven.sterck]

ok, yes, that's what I mean with 'by position'. I don't think what you ask is possible with any of the existing tools. Splitting the BAM, as you mentioned, is likely the most easy way to get to it.

[vitor.rca]

OK, thank you. It takes a long time to split a BAM with bamtools split, I mean, more time than I was anticipating for a BAM already sorted by position. So I was hoping for an alternative, but it seems like there's no reason for someone to implement such sorting scheme, since no tool requires it, except for the UBU sam-xlate tool which I'm trying to use (https://github.com/mozack/ubu/wiki).

Answer 1

[Pierre Lindenbaum]

I wrote http://lindenb.github.io/jvarkit/SortSamRefName.html

[lieven.sterck]

damn, should have known you would have code at hand for that Pierre Lindenbaum . :-)

thx.