pdb ftp path updated

flatironinstitute · Jan 26, 2021 · e1d8ffd · e1d8ffd
1 parent 601d9fd
commit e1d8ffd
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 2 deletions.
diff --git a/preprocessing/create_nrPDB_GO_annot.py b/preprocessing/create_nrPDB_GO_annot.py
@@ -138,16 +138,25 @@ def write_output_files(fname, pdb2go, go2info, pdb2seq):
  onts = ['molecular_function', 'biological_process', 'cellular_component']
  selected_goterms = {ont: set() for ont in onts}
  selected_proteins = set()
+ for goterm in go2info:
+ prots = go2info[goterm]['pdb_chains']
+ num = len(prots)
+ namespace = go2info[goterm]['ont']
+ if num > 49 and num <= 5000:
+ selected_goterms[namespace].add(goterm)
+ selected_proteins = selected_proteins.union(prots)
+ """
  for chain in pdb2go:
  goterms = set(pdb2go[chain]['goterms'])
  if len(goterms) > 2 and chain in pdb2seq:
  for goterm in goterms:
  prots = go2info[goterm]['pdb_chains']
  num = len(prots)
  namespace = go2info[goterm]['ont']
- if num > 49 and num < 5000:
+ if num > 19 and num <= 5000:
  selected_goterms[namespace].add(goterm)
  selected_proteins = selected_proteins.union(prots)
+ """
 
  selected_goterms_list = {ont: list(selected_goterms[ont]) for ont in onts}
  selected_gonames_list = {ont: [go2info[goterm]['goname'] for goterm in selected_goterms_list[ont]] for ont in onts}

diff --git a/preprocessing/data_collection.sh b/preprocessing/data_collection.sh
@@ -17,7 +17,7 @@ printf "\n\n DOWNLOADING PDB SEQRES SEQUENCES...\n"
 wget ftp://ftp.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt.gz -O $DATA_DIR/pdb_seqres.txt.gz
 
 printf "\n\n DOWNLOADING PDB CLUSTERS...\n"
-wget ftp://resources.rcsb.org/sequence/clusters/bc-$SEQ_SIM.out -O $DATA_DIR/bc-$SEQ_SIM.out
+wget https://cdn.rcsb.org/resources/sequence/clusters/bc-$SEQ_SIM.out -O $DATA_DIR/bc-$SEQ_SIM.out
 
 printf "\n\n DOWNLOADING GO HIERARCHY...\n"
 wget http://purl.obolibrary.org/obo/go/go-basic.obo -O $DATA_DIR/go-basic.obo