Skip to content

Replace Bio::DB::Fasta with faidx_xs

Marek Szuba requested to merge github/fork/willmclaren/feature/faidx into master

Created by: willmclaren

Benefits:

  • can index bgzipped FASTA files (though note Ensembl distributes gzipped, must be converted to bgzipped before use). Human toplevel file is ~1.1g bgzipped, 40g uncompressed.
  • indexes much more robust, won't tolerate corrupted/partial index like Bio::DB::Fasta
  • around 10x faster retrieving sequence

Requires:

The .travis.yml install of htslib and faidx_xs is non-ideal as it avoids using "make install" by compiling the module into the home directory. Much better if you have sudo access to do "make install" properly for each.

Merge request reports