Supporting energize_docking runs in condor framework

Author: samgelman

code/condor.py

@@ -169,6 +169,17 @@ def load_lines(fn):
 
 
 def prep_energize(args):
+    """
+    Prepare a condor run for calculating Rosetta energies
+    """
+
+    # supports original energies (energize.py) or docking for GB1 docking energies (gb1_docking.py)
+    if args.run_type == "energize":
+        pyscript = "energize.py"
+    elif args.run_type == "energize_docking":
+        pyscript = "gb1_docking.py"
+    else:
+        raise ValueError("Invalid run type: {}".format(args.run_type))
 
     out_dir = join("output", "htcondor_runs", get_run_dir_name(args.run_name))
     os.makedirs(out_dir)
@@ -190,6 +201,8 @@ def prep_energize(args):
     with open(join(out_dir, "env_vars.txt"), "w") as f:
         f.write("export GITHUB_TAG={}\n".format(args.github_tag))
         f.write("export NUM_JOBS={}\n".format(num_jobs))
+        # additional environment variables to handle the different types of runs
+        f.write("export PYSCRIPT={}\n".format(pyscript))
 
     # prepare the additional data files
     additional_files = prep_additional_data_files(args.additional_data_files, out_dir)
@@ -204,7 +217,7 @@ def prep_energize(args):
     shutil.copy("htcondor/templates/run.sh", out_dir)
     shutil.copy("htcondor/templates/pass.txt", out_dir)
 
-    # copy over the energize args and rename to standard filename
+    # copy over energize args and rename to standard filename
     shutil.copyfile(args.energize_args_fn, join(out_dir, "energize_args.txt"))
 
     # create output directories where jobs will place their outputs
@@ -388,7 +401,7 @@ def main(args):
                         help="prepare or energize",
                         type=str,
                         default="energize",
-                        choices=["prepare", "energize"])
+                        choices=["prepare", "energize", "energize_docking"])
 
     parser.add_argument("--run_name",
                         help="name for this condor run, used for log directory",

docking_args/gb1_1fcc_condor_set_1.txt

@@ -0,0 +1,10 @@
+--rosetta_main_dir
+rosetta_minimal
+--pdb_dir
+pdb_files/prepared_pdb_files
+--chain
+C
+--num_structs
+1
+--log_dir_base
+output/energize_outputs

htcondor/templates/energize.sub

@@ -22,7 +22,7 @@ request_cpus = 1
 request_memory = 2GB
 request_disk = 20GB
 
-environment = "CLUSTER=$(Cluster) PROCESS=$(Process) RUNNINGON=$$(Name) GITHUB_TAG=$ENV(GITHUB_TAG)"
+environment = "CLUSTER=$(Cluster) PROCESS=$(Process) RUNNINGON=$$(Name) GITHUB_TAG=$ENV(GITHUB_TAG) PYSCRIPT=$ENV(PYSCRIPT)"
 
 # when submitting more than 10k jobs this will prevent the scheduler from getting overloaded
 max_idle = 1000

htcondor/templates/run.sh

@@ -58,7 +58,6 @@ tar -xzf rosettafy_env_v0.7.11.tar.gz -C rosettafy_env
 . rosettafy_env/bin/activate
 rm rosettafy_env_v0.7.11.tar.gz
 
-
 # decrypt
 # note this is done AFTER setting up the Python environment because it requires
 # the openssl version inside the environment
@@ -79,6 +78,5 @@ then
 fi
 
 # launch our python run script with argument file number
-echo "Launching energize.py"
-python3 code/energize.py @energize_args.txt --variants_fn="${PROCESS}.txt" --cluster="$CLUSTER" --process="$PROCESS" --commit_id="$GITHUB_TAG"
-
+echo "Launching ${PYSCRIPT}"
+python3 code/${PYSCRIPT} @energize_args.txt --variants_fn="${PROCESS}.txt" --cluster="$CLUSTER" --process="$PROCESS" --commit_id="$GITHUB_TAG"