feat: do everything with the foldseek folder

backend/src/api/similar.py

@@ -4,15 +4,18 @@
  pdb_file_name,
  revert_pdb_filename,
 )
-from ..foldseek import easy_search, create_db
+from ..foldseek import easy_search, external_db
 
 router = APIRouter()
 
 
+PDB = external_db("PDB", "foldseek/dbs/pdb")
+
+
 @router.get("/similar-pdb/{protein_name:str}", response_model=list[SimilarProtein])
 def get_similar_pdb(protein_name: str):
+ global PDB
  query_name = pdb_file_name(protein_name)
- PDB = create_db("PDB", "pdb")
  similar = easy_search(query_name, PDB, out_format="target,prob")
  return [SimilarProtein(name=s[0].split(".")[0], prob=s[1]) for s in similar]
 

backend/src/foldseek.py

@@ -23,7 +23,7 @@ class GenerateDirName:
  def __enter__(self):
  global active_caches
  active_caches += 1
- self.temp_dir = ".foldseek_cache_" + str(active_caches)
+ self.temp_dir = "foldseek/temp_dir_" + str(active_caches)
  return self.temp_dir
 
  def __exit__(self, *args):
@@ -93,6 +93,30 @@ def easy_search(
  return to_columnar_array(parsed_output) if columnar else parsed_output
 
 
+def external_db(
+ external_db_name: str,
+ db_name: str,
+ foldseek_executable="foldseek/bin/foldseek",
+ print_stdout=False,
+):
+ if external_db_name not in EXTERNAL_DATABASES:
+ raise Exception(f"Directory {external_db_name} not found")
+
+ with GenerateDirName() as temp_dir:
+ try:
+ bash_cmd(f"ls {db_name}")
+ except Exception:
+ if dir not in EXTERNAL_DATABASES:
+ cmd = f"{foldseek_executable} createdb {external_db_name} {db_name}"
+ else:
+ cmd = f"{foldseek_executable} databases {external_db_name} {db_name} {temp_dir}"
+ stdout = bash_cmd(cmd)
+ if print_stdout:
+ print(stdout)
+
+ return db_name
+
+
 def create_db(
  dir: str,
  db_name: str,