More comprehensive Pfam ID mapping

genome · Apr 21, 2017 · 2ad4a69 · 2ad4a69
1 parent ee19a7f
commit 2ad4a69
Show file tree

Hide file tree

Showing 8 changed files with 8 additions and 8 deletions.
diff --git a/.gitignore b/.gitignore
@@ -2,3 +2,4 @@
 doc/_build/**
 doc/_static/**
 test/DLG1-BRAF/*
+submit_to_pip.sh
diff --git a/MANIFEST.in b/MANIFEST.in
@@ -1,6 +1,6 @@
 include README.md
 include agfusion/data/agfusion.db.gz
-include agfusion/data/042017_pdb_pfam_mapping.txt.gz
+include agfusion/data/042117_pfamA.txt.gz
 include bin/agfusion
 include test/test.py
 include test/test_mouse.sh

diff --git a/agfusion/data/042017_pdb_pfam_mapping.txt.gz b/agfusion/data/042017_pdb_pfam_mapping.txt.gz
diff --git a/agfusion/data/042117_pfamA.txt.gz b/agfusion/data/042117_pfamA.txt.gz
diff --git a/agfusion/data/agfusion.db.gz b/agfusion/data/agfusion.db.gz
diff --git a/agfusion/data/build_db.sh b/agfusion/data/build_db.sh
@@ -1,3 +1,4 @@
+curl ftp://ftp.ebi.ac.uk/pub/databases/Pfam/current_release/database_files//pfamA.txt.gz > 042117_pfamA.txt.gz
 
 agfusion database --database agfusion.db --build homo_sapiens_core_84_38
 

diff --git a/agfusion/database.py b/agfusion/database.py
@@ -122,17 +122,15 @@ def __init__(self, database, build, server):
         file_path = os.path.join(
             os.path.split(__file__)[0],
             'data',
-            '042017_pdb_pfam_mapping.txt.gz'
+            '042117_pfamA.txt.gz'
         )
 
         for line in gzip.open(file_path, 'rb'):
-            if line.find('PDB_ID')!=-1:
-                next
             line = line.rstrip().split('\t')
 
-            pfam_id = line[4].split('.')[0]
-            pfam_name = line[5]
-            pfam_desc = line[6]
+            pfam_id = line[0]
+            pfam_name = line[1]
+            pfam_desc = line[3]
 
             self.pfam_mapping[pfam_id] = {
                 'name': pfam_name,

diff --git a/test/test_human.sh b/test/test_human.sh
@@ -7,7 +7,7 @@
 #  --genome GRCh37
 
 
-agfusion \
+agfusion annotate \
   --gene5prime ENSG00000066468 \
   --gene3prime ENSG00000197959 \
   --junction5prime 121485532 \
Original file line number	Diff line number	Diff line change
		@@ -1,3 +1,4 @@
		curl ftp://ftp.ebi.ac.uk/pub/databases/Pfam/current_release/database_files//pfamA.txt.gz > 042117_pfamA.txt.gz

		agfusion database --database agfusion.db --build homo_sapiens_core_84_38

Expand Down