Skip to content

After installing dependencies for running the protein trees pipeline to google cloud platform, a set of shell scripts were developed. These are those.

Notifications You must be signed in to change notification settings

carlacummins/compara-cloud-dependencies

Repository files navigation

compara-cloud-dependencies

After installing dependencies for running the protein trees pipeline to google cloud platform, a set of shell scripts were developed. These are those.

These instructions are an amalgamation of the following:

Steps

  1. source envs.sh
  2. run shell scripts in order:
    • 1.linuxbrew_basic_dependencies.sh
    • 2.linuxbrew_compara_base.sh
    • 3.linuxbrew_compara_proteintrees.sh

Notes / Known Issues

xz

  • xz was failing make check tests when installing through brew install ensembl/cask/basic-dependencies
    • solution was to manually install xz in the correct location: ./configure --prefix=$ENSEMBL_LINUXBREW_DIR/Cellar/xz/5.2.5 && make && make install
    • brew link xz

percona-client

  • percona-client install failing: $ENSEMBL_LINUXBREW_DIR/include/sys/types.h:33:9: error: unknown type name '__u_char' typedef __u_char u_char;
    • tried editing formula ($ENSEMBL_LINUXBREW_DIR/Library/Taps/ensembl/homebrew-external/percona-client.rb) to include new CFLAGs: -Du_char="unsigned char"
    • these needed to be included in both the cmake and make commands : not successful
    • solution was to remove dependencies (and hope it doesn't bite later):
      • edit /apps/compara/linuxbrew/Library/Taps/ensembl/homebrew-cask/yml/perl-clibs.yml and remove percona, kent (as it has percona as a dependency)
      • regenerate the cask: bin/create_cask --verbose --config yml/perl-clibs.yml --dir .
      • remove cached files: ~/.cache/Homebrew/downloads/<SHA256>--<dep name>.rb
      • brew install ensembl/cask/perl-clibs again

graphviz / bison

hcluster

  • trouble with hcluster: formula is set to use svn to download v0.5.0
    • brew install svn is breaking down due to krb5 error
    • first try: edit formula to grab a v0.5.1 I found on github: https://github.com/douglasgscofield/hcluster
      • can then update the download strategy in the formula to :using => :git

harfbuzz

DBD::mysql

  • DBD::mysql is a nightmare to install!!
    • in the end, some environment variables (see envs.sh) to set up mysql connection credentials helped
    • needed to set export LD_LIBRARY_PATH=/usr/lib:/usr/lib64:/usr/lib64/mysql so that all the libs were found
    • cpanm DBD::mysql --force had to be used because of testing failures

About

After installing dependencies for running the protein trees pipeline to google cloud platform, a set of shell scripts were developed. These are those.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages