Download database now extra step

Have to download AGFusion database. This reduces the python package size

MANIFEST.in

@@ -1,6 +1,4 @@
 include README.md
-include agfusion/data/agfusion.db.gz
-include agfusion/data/042117_pfamA.txt.gz
 include bin/agfusion
 include test/test.py
 include test/test_mouse.sh

README.md

@@ -150,12 +150,6 @@ agfusion annotate \
 ![alt tag](https://github.com/murphycj/AGFusion/blob/master/doc/ENSMUST00000064477-ENSMUST00000002487-rescale.png)
 ![alt tag](https://github.com/murphycj/AGFusion/blob/master/doc/ENSMUST00000122054-ENSMUST00000070330-rescale.png)
 
-# Advanced Usage
-
-### Building your own database
-
-Under construction...
-
 # Installation
 
 First you need to install pyensembl (and the other dependencies listed at the bottom) and download the reference genome you will use by running one of the following.
@@ -184,6 +178,12 @@ Then you can install AGFusion via the following:
 pip install agfusion
 ```
 
+Finally, download the AGFusion database:
+
+```
+agfusion download
+```
+
 # Dependencies
 
 - python 2.7.8

agfusion/_version.py

@@ -1 +1 @@
-__version__ = "0.138"
+__version__ = "0.139"

agfusion/cli.py

@@ -11,6 +11,25 @@
 from agfusion import exceptions
 import pyensembl
 
+def downloaddb(args):
+
+    import gzip
+    import urllib
+
+    if not os.path.exists(args.dir):
+        os.mkdir(args.dir)
+
+    file_path = os.path.join(
+        args.dir,
+        'agfusion.db.gz')
+
+    urllib.urlretrieve(
+        "https://raw.githubusercontent.com/murphycj/AGFusionDB/master/agfusion.db.gz",
+        file_path)
+
+    with gzip.open(file_path, 'rb') as f_in, file(file_path.replace('.gz',''), 'w') as f_out:
+        shutil.copyfileobj(f_in, f_out)
+
 def annotate(gene5prime,junction5prime,gene3prime,junction3prime,
              outdir,colors,rename,scale,db,pyensembl_data,args):
 
@@ -190,12 +209,11 @@ def add_common_flags(parser):
         help='(Optional) Do not label domains.'
     )
     parser.add_argument(
-        '--db',
+        '--dbpath',
         type=str,
-        default=None,
+        default=os.path.join(os.path.expanduser('~'),'.agfusion'),
         required=False,
-        help='(Optional) The SQLite3 database. Defaults to using the ' +
-             'database provided by the package.'
+        help='(Optional) Path to where the AGFusion databse is located (default: ' + os.path.join(os.path.expanduser('~'),'.agfusion') + ')'
     )
     parser.add_argument(
         '--debug',
@@ -287,24 +305,36 @@ def main():
     )
     add_common_flags(batch_parser)
 
-    # download database parser
+    # download database
 
-    database_parser = subparsers.add_parser('database', help='Download database for a reference genome.')
+    database_parser = subparsers.add_parser('download', help='Download database for a reference genome.')
     database_parser.add_argument(
+        '--dir',
+        type=str,
+        required=False,
+        default=os.path.join(os.path.expanduser('~'),'.agfusion'),
+        help='(Optional) Directory to the database will be downloaded to (default: $HOME/.agfusion/)'
+    )
+    args = parser.parse_args()
+
+    # build database parser
+
+    build_database_parser = subparsers.add_parser('build', help='Build database for a reference genome.')
+    build_database_parser.add_argument(
         '--database',
         type=str,
         required=True,
         help='Path to the database file (e.g. agfusion.db)'
     )
-    database_parser.add_argument(
+    build_database_parser.add_argument(
         '--build',
         type=str,
         required=True,
         help='homo_sapiens_core_84_38 (for GRCh38), ' +
         'homo_sapiens_core_75_37 (for GRCh37), or ' +
         'mus_musculus_core_84_38 (for GRCm38)'
     )
-    database_parser.add_argument(
+    build_database_parser.add_argument(
         '--server',
         type=str,
         required=False,
@@ -313,25 +343,22 @@ def main():
     )
     args = parser.parse_args()
 
-    if args.subparser_name == 'database':
+    if args.subparser_name == 'build':
         builddb(args)
+    elif args.subparser_name == 'download':
+        downloaddb(args)
     else:
         if not os.path.exists(args.out):
             os.mkdir(args.out)
 
         # if user does not specify a sqlite database then use the one provided
         # by the package
 
-        if args.db is None:
-            file_path = os.path.join(
-                os.path.split(__file__)[0],
-                'data',
-                'agfusion.db'
-            )
-
-            db = agfusion.AGFusionDB(file_path,debug=args.debug)
-        else:
-            db = agfusion.AGFusionDB(args.db,debug=args.debug)
+        file_path = os.path.join(
+            args.dbpath,
+            'agfusion.db'
+        )
+        db = agfusion.AGFusionDB(file_path,debug=args.debug)
 
         # get the pyensembl data
 

setup.py

@@ -1,24 +1,9 @@
 import os
-import site
-import gzip
-from setuptools import setup, find_packages
-from setuptools.command.install import install
 import re
 import shutil
+import site
 
-class CustomInstall(install):
-    def run(self):
-        install.run(self)
-
-        file_path = os.path.join(
-            site.getsitepackages()[0],
-            'agfusion',
-            'data',
-            'agfusion.db.gz'
-        )
-
-        with gzip.open(file_path, 'rb') as f_in, file(file_path.replace('.gz',''), 'w') as f_out:
-            shutil.copyfileobj(f_in, f_out)
+from setuptools import setup, find_packages
 
 VERSIONFILE = "agfusion/_version.py"
 verstrline = open(VERSIONFILE, "rt").read()
@@ -35,15 +20,14 @@ def run(self):
     version=verstr,
     name='agfusion',
     packages=find_packages(),
-    description="Python package providing that can visualize different annotations of a gene fusion.",
+    description="Python package to annotate and visualize gene fusions.",
     author='Charles Murphy',
     author_email='[email protected]',
     license='MIT',
     url='https://github.com/murphycj/AGFusion',
     long_description=README,
     include_package_data=True,
     scripts=['bin/agfusion'],
-    cmdclass={'install': CustomInstall},
     install_requires=[
         'pyensembl>=0.9.5',
         'matplotlib>=1.5.0',

test/test.py

@@ -1,11 +1,12 @@
+from os.path import join, expanduser
 import unittest
 import agfusion
 from agfusion import utils
 import pyensembl
 from Bio import SeqIO
 
 data = pyensembl.EnsemblRelease(84,'mouse')
-db = agfusion.AGFusionDB(utils.agfusion_db)
+db = agfusion.AGFusionDB(join(expanduser('~'),'.agfusion','agfusion.db'))
 db.build = 'mus_musculus_core_84_38'
 
 class TestSequencePrediction(unittest.TestCase):