greenelab / annotation-refinery Goto Github PK

This is to follow a loose-coupling model, so that certain parts of the code can be imported and called from anywhere.

This issue was imported from the old repository in bitbucket (https://bitbucket.org/greenelab/annotation-refinery/issues).

This issue was imported from the old repository in bitbucket (https://bitbucket.org/greenelab/annotation-refinery/issues).

This was brought up in #8.

The ftp://ftp.geneontology.org/go/gene-associations/goa_uniprot_all_noiea.gaf.gz URL was being used as one of the GO ASSOC_FILE_URLS for all of the species, so we thought we could abstract it to the main_config.ini file. It can have a name like COMMON_GO_ASSOC_URLS or something similar, and the code in download_files.py and process_go.py will need to get it from this new location, instead of its old location in the .ini files.

This sounds like it is more of an issue in Tribe than the annotation-refinery. Cross-linking with Tribe issue:
greenelab/tribe#5

This would be used in the process_go.py, process_kegg.py and process_do.py files.

As per Matt's suggestion, it could look something like this:

def filename_from_url(self, folder, urlstr): 
    return os.path.join(folder, 'DO', os.path.basename(urlsplit(urlstr).path))

This issue was imported from the old repository in bitbucket (https://bitbucket.org/greenelab/annotation-refinery/issues).

This is a very minor issue:

When running cron_update_public_tribe.sh initially on my desktop, I noticed that line 5:
rm -r download_files/
complained that download_files doesn't exist. It can be fixed by either:

if [ -e download_files ]; then
    rm -r download_files
fi

or simply:
rm -rf download_files

(Note that the trailing / in download_files/ is removed too, because download_files should be deleted regardless of whether it is a directory or not.)

I found a few typos in README.rst:

1. Line 19: amon should be among.

2. Line 33: loaded to probably should be uploaded to, downloaded to should be downloaded from.

3. Line 43: this files should be this file.

This affects the download_from_url() function in utils.py. This is a documented issue in Python, which persists in the newest version (v2.7.13), at the time this issue was created (http://bugs.python.org/issue27973).

This is an older issue, copied from https://bitbucket.org/rzelayafavila/annotation-refinery/issues