Just following up on this to suggest that the ergonomics of this are less than ideal. I'm blissfully unaware of the constraints that samtools developers face, but maybe a solution would be to create the REF_CACHE stuff from a given fasta in a tmp dir (or in memory or in $REF_CACHE) if it's needed and not present.
The long version is that I am trying to index a cram file for which I have the reference fasta. I am getting messages about REF_CACHE. If samtools needs the reference in order to index, how can I specify that? I'm sure I've indexed without a reference but I am hitting that in this case. after some time it says: [W::cram_populate_ref] Creating reference cache directory /root/.cache/hts-ref it also tells me: "This may become large; see the samtools(1) manual page REF_CACHE discussion" The coverage of what REF_CACHE and REF_PATH are is not covered in enough detail to be able to use them in the man page (http://www.htslib.org/doc/samtools.html) though there is more detail in the workflow page (http://www.htslib.org/workflow/). In case anyone else encounters this, in order to get this working without having it autodownload to /root/.cache, I had to do: wget -q https://raw.githubusercontent.com/samtools/samtools/develop/misc/seq_cache_populate.pl perl seq_cache_populate.pl -root $(pwd)/cache ${fasta} export REF_PATH=$(pwd)/cache/%2s/%2s/%s:http://www.ebi.ac.uk/ena/cram/md5/%s export REF_CACHE=xx -Brent On Wed, Feb 28, 2018 at 10:40 AM Brent Pedersen <bpede...@gmail.com> wrote: > > On Thu, Feb 8, 2018 at 11:13 AM, James Bonfield <j...@sanger.ac.uk> wrote: > > On Thu, Feb 08, 2018 at 09:32:00AM -0700, Brent Pedersen wrote: > >> I've been working more with crams lately and in some cases, it seems > >> the default behavior of htslib is to automatically start downloading > >> reference files locally even though I have not set REF_CACHE or > >> REF_PATH. This is deep in a complex pipeline behind several layers of > >> abstraction, but I think this is the default. > > > > If you don't set them it'll automatically default to using the EBI for > > REF_PATH and ~/.cache/hts-seqs (IIRC) for REF_CACHE. > > > > If you really wish to totally disable them, keep REF_CACHE clear and > > set REF_PATH to somewhere with no files, eg /foo, and it'll fail the > > search. > > > > It should then fail. > > > > James > > > > -- > > James Bonfield (j...@sanger.ac.uk) > > The Sanger Institute, Hinxton, Cambs, CB10 1SA > > > > > > -- > > The Wellcome Sanger Institute is operated by Genome Research > > Limited, a charity registered in England with number 1021457 and a > > company registered in England with number 2742969, whose registered > > office is 215 Euston Road, London, NW1 2BE. > > > Thanks for the reply. It seems a better solution would be to fail > unless REF_CACHE or REF_PATH are set. > I'm having scenarios where a user uses the wrong reference and > samtools detects this and starts downloading with message: "Creating > reference cache directory $HOME/.cache/hts-ref" > > I never want to have stuff download automatically. Even setting > REF_PATH=xx , it will start downloading if I specify the wrong > reference. > -Brent _______________________________________________ Samtools-help mailing list Samtools-help@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/samtools-help
