Besides Django that is. Both packages where installed using Python's pip(1) command.

The user running YeastPhenome needs write acces to a $HOME/libChEBI directory. When in production this is likely the Apache user. On CentOS mkdir /var/www/libChEBI; chown apache.apache /var/www/libChEBI is needed. As of this writing libChEBIpy didn't seem to have a way to alter this location.

yeastphenome/settings.py-template needs to be copied to yeastphenome/settings.py and edited to suit your needs. Currently there are a few added items that isn't a Django standard.

Link to directory containing directoris of PubMed IDs to be offered as zip files.

The prefix to use when downloading a zip file.

Path to a readme file in be included in the downloaded zip file.

Rather then storing it in the database MEDLINE formatted information is stored it it's raw text format in the directory sepcified here. This must also be accessed via the url .../static/MEDLINE/*pmid*.txt URL. So the default is probably what you want.

The migrate option doesn't seem to work in one step when using PostgreSQL, so you will need to comment out part of the INSTALLED_APPS section the first time you run it:

INSTALLED_APPS = (
    # For some odd reason when we first migrate with postgres we need
    # to do only these three first.
    'django.contrib.auth',
    'django.contrib.sites',
    'django.contrib.contenttypes',

    # # Then add the rest when the first three are successful.
    # 'django.contrib.admin',
    # 'django.contrib.sessions',
    # 'django.contrib.messages',
    # 'django.contrib.staticfiles',
    # 'google_analytics',

    # 'papers',
    # 'phenotypes',
    # 'conditions',
)

Then migrate the database. After that, uncomment out the bottom part of the block and migrate again.

As curators use the admin site to curate it is important to keep the static/admin up with the current version of Django you are using. So when using something like WSGI make sure that it points to the code in the python library.