[Bioc-devel] GenomeInfoDb::Seqinfo(genome) broken?

I think this is because the NCBI server switched to https (via a
redirect that I guess the R url() connection fails to follow). The
reason rtracklayer still works is that it's only querying UCSC.
GenomeInfoDb also queries NCBI to get the mappings to the NCBI
seqlevels. Does that really need to happen when only getting the
Seqinfo?


On Thu, Nov 3, 2016 at 5:13 AM, Raymond Cavalcante <rcavalca at umich.edu> wrote:

Hello,

Sometime yesterday calls like GenomeInfoDb::Seqinfo(genome = 'hg19') stopped working with the error:

Error in file(file, "rt") : cannot open the connection

From the documentation, that call relies on fetchExtendedChromInfoFromUCSC() and requires an internet connection, which I had and continue to have. I'm not really sure how to deal with this problem because the goldenPath link still works (http://hgdownload.cse.ucsc.edu/goldenpath/hg19/database/chromInfo.txt.gz <http://hgdownload.cse.ucsc.edu/goldenpath/hg19/database/chromInfo.txt.gz>), so something else is broken...

Oddly, calls to rtracklayer::import.bed() that specify a genome work. I don't have any BSgenome packages installed where I'm running it, and from the documentation for genome, "An attempt will be made to derive the ?seqinfo? on the return value using either an installed BSgenome package or UCSC, if network access is available." So I would guess that rtracklayer::import.bed() would use the same fetchExtendedChromInfoFromUCSC()...?

On a related note, is there a non-BSgenome package that has the chromosome length / seqinfo information that doesn't require an internet connection (other than to download the package)? BSgenome is too large to require of users just for chromosome lengths. The org.db packages have chromosome lengths, but only with respect to one genome version for that organism, and from the documentation it isn't clear which version.

Thanks,
Raymond Cavalcante
        [[alternative HTML version deleted]]