Tweaks to allow loading old data

also some changes to BindingPMF API

AlGDock/BindingPMF.py

@@ -18,7 +18,6 @@
 from collections import OrderedDict
 
 from AlGDock import dictionary_tools
-from AlGDock import path_tools
 
 import MMTK
 import MMTK.Units
@@ -36,11 +35,6 @@
 import multiprocessing
 from multiprocessing import Process
 
-try:
-  import requests  # for downloading additional files
-except:
-  print '  no requests module for downloading additional files'
-
 # For profiling. Unnecessary for normal execution.
 # from memory_profiler import profile
 
@@ -78,13 +72,13 @@ def HMStime(s):
 
 class BPMF:
   def __init__(self, **kwargs):
-    """Parses the input arguments and runs the requested CD calculation"""
+    """Parses the input arguments and runs the requested calculation"""
 
     #         mod_path = os.path.join(os.path.dirname(a.__file__), 'BindingPMF.py')
     #         print """###########
     # # AlGDock #
     # ###########
-    # Molecular CD with adaptively scaled alchemical interaction grids
+    # Molecular docking with adaptively scaled alchemical interaction grids
     #
     # in {0}
     # last modified {1}
@@ -100,6 +94,8 @@ def __init__(self, **kwargs):
     self.data['CD'] = SimulationData(self.args.dir['CD'], 'CD', \
       self.args.params['CD']['pose'])
 
+    if not 'max_time' in kwargs.keys():
+      kwargs['max_time'] = None
     if not 'run_type' in kwargs.keys():
       kwargs['run_type'] = None
 
@@ -117,7 +113,7 @@ def __init__(self, **kwargs):
       print '\nfor %s:' % p
       print dictionary_tools.dict_view(self.args.params[p])[:-1]
 
-    self._run(kwargs['run_type'])
+    self.run(kwargs['run_type'])
 
   def _setup(self):
     """Creates an MMTK InfiniteUniverse and adds the ligand"""
@@ -249,13 +245,6 @@ def _setup(self):
       self.top.universe.setConfiguration(
         Configuration(self.top.universe, confs[-1]))
 
-    # Samplers may accept the following options:
-    # steps - number of MD steps
-    # T - temperature in K
-    # delta_t - MD time step
-    # normalize - normalizes configurations
-    # adapt - uses an adaptive time step
-
     from AlGDock.simulation_iterator import SimulationIterator
     self.iterator = SimulationIterator(self.args, self.top, self.system)
 
@@ -269,7 +258,7 @@ def _setup(self):
     if self.args.random_seed > 0:
       np.random.seed(self.args.random_seed)
 
-  def _run(self, run_type):
+  def run(self, run_type):
     from AlGDock.postprocessing import Postprocessing
 
     self.log.recordStart('run')
@@ -484,7 +473,7 @@ def calc_f_L(self, readOnly=False, do_solvation=True, redo=False):
       for BC_Es_state in self.data['BC'].Es:
         BC_Es.append(BC_Es_state[fromCycle:toCycle])
       (u_kln, N_k) = self._u_kln(BC_Es, self.data['BC'].protocol)
-      MBAR = self._run_MBAR(u_kln, N_k)[0]
+      MBAR = self.run_MBAR(u_kln, N_k)[0]
       self.f_L['BC_MBAR'].append(MBAR)
 
       # Average acceptance probabilities
@@ -714,7 +703,7 @@ def set_updated_to_True(updated, start_string, quiet=False):
 
       # Use MBAR for the grid scaling free energy estimate
       (u_kln, N_k) = self._u_kln(CD_Es, self.data['CD'].protocol)
-      MBAR = self._run_MBAR(u_kln, N_k)[0]
+      MBAR = self.run_MBAR(u_kln, N_k)[0]
       self.f_RL['grid_MBAR'][c] = MBAR
       updated = set_updated_to_True(updated,
                                     start_string,
@@ -1144,8 +1133,6 @@ def configuration_energies(self, minimize=False, max_confs=None):
       confs = confs[:max_confs]
 
     # Implicit solvent energies
-    self._load_programs(self.args.params['CD']['phases'])
-
     self.data['CD'].confs['starting_poses'] = None
     from AlGDock.postprocessing import Postprocessing
     pp_complete = Postprocessing(self.args, self.log, self.top, self.top_RL, self.system, self.data, self.save).run([('original', 0, 0, 'R')])
@@ -1380,7 +1367,7 @@ def _insert_CD_state(self, alpha, clear=True):
       u_kln[k, -1, :] = u_kln_n[k, 0, :]
 
     # Determine SIR weights
-    weights = self._run_MBAR(u_kln, N_k, augmented=True)[1][:, -1]
+    weights = self.run_MBAR(u_kln, N_k, augmented=True)[1][:, -1]
     weights = weights / sum(weights)
 
     # Resampling
@@ -1692,7 +1679,7 @@ def _checkedMinimizer(self, confs):
       ', '.join(['%.2f'%e for e in energies[:10]]))
     return confs, energies
 
-  def _run_MBAR(self, u_kln, N_k, augmented=False):
+  def run_MBAR(self, u_kln, N_k, augmented=False):
     """
     Estimates the free energy of a transition using BAR and MBAR
     """

AlGDock/path_tools.py

@@ -5,8 +5,10 @@
 
 ### Google drive downloader from
 # http://stackoverflow.com/questions/25010369/wget-curl-large-file-from-google-drive
-import requests
-
+try:
+  import requests  # for downloading additional files
+except:
+  print '  no requests module for downloading additional files'
 
 def download_file_from_google_drive(id, destination):
   URL = "https://docs.google.com/uc?export=download"

AlGDock/postprocessing.py

@@ -78,7 +78,13 @@ def __init__(self, args, log, top, top_RL, system, data, save):
       self.elsize = self._get_elsize()
 
   def _load_programs(self, phases):
-    # Find the necessary programs, downloading them if necessary
+    """Locate the necessary external programs, downloading them if necessary
+
+    Parameters
+    ----------
+    phases : list of str
+      The names of the phases that will be used
+    """
     programs = []
     for phase in phases:
       for (prefix,program) in [('NAMD','namd'), \
@@ -103,13 +109,24 @@ def run(self,
                   ('CD',   -1,-1,'L'), ('CD',-1,-1,'RL')],
       phases=None,
       readOnly=False, redo_CD=False, debug=DEBUG):
-    """
-    Obtains the NAMD energies of all the conditions using all the phases.
-    Saves both MMTK and NAMD energies after NAMD energies are estimated.
+    """Obtains and stores energies of all conditions in all phases
 
-    state == -1 means the last state
-    cycle == -1 means all cycles
+    Energies are stored in self.data
 
+    Parameters
+    ----------
+    conditions : list of tuples
+      Each tuple contains the process, the thermodynamic state index,
+      the simulation cycle, and moeity.
+
+      state == -1 means the last state
+      cycle == -1 means all cycles
+
+    phases : list of str
+      The names of the phases that will be considered
+    readOnly : bool
+      If True, no calculations will be performed. The data will simply be
+      loaded into memory.
     """
     # Clear evaluators to save memory
     self.system.clear_evaluators()
@@ -317,6 +334,8 @@ def run(self,
       return len(incomplete) == len(results)
 
   def _energy_worker(self, input, output, time_per_snap):
+    """Performs energy calculation for one process
+    """
     for args in iter(input.get, 'STOP'):
       (confs, moiety, phase, traj_FN, outputname, debug, reference) = args
       (p, state, c, label) = reference
@@ -358,8 +377,7 @@ def _NAMD_Energy(self,
                    outputname,
                    debug=DEBUG,
                    reference=None):
-    """
-    Uses NAMD to calculate the energy of a set of configurations
+    """Uses NAMD to calculate the energy of a set of configurations.
     Units are the MMTK standard, kJ/mol
     """
     # NAMD ENERGY FIELDS:
@@ -556,8 +574,10 @@ def _sander_Energy(self, confs, moiety, phase, AMBER_mdcrd_FN, \
     # 5. EPB 6. 1-4 VWD 7. 1-4 EEL 8. RESTRAINT 9. ECAVITY 10. EDISPER
 
   def _get_elsize(self):
-    # Calculates the electrostatic size of the receptor for ALPB calculations
-    # Writes the coordinates in AMBER format
+    """Calculates the electrostatic size of the receptor for ALPB calculations
+
+    Writes the coordinates in AMBER format
+    """
     if hasattr(self, 'elsize'):
       return
 
@@ -623,9 +643,7 @@ def _gbnsr6_Energy(self,
                      outputname,
                      debug=DEBUG,
                      reference=None):
-    """
-    Uses gbnsr6 (part of AmberTools)
-    to calculate the energy of a set of configurations
+    """Uses gbnsr6 (part of AmberTools) to calculate the energy of a set of configurations
     """
     # Prepare configurations for writing to crd file
     factor = 1.0 / MMTK.Units.Ang
@@ -731,9 +749,8 @@ def _APBS_Energy(self,
                    outputname,
                    debug=DEBUG,
                    reference=None):
-    """
-    Uses APBS to calculate the solvation energy of a set of configurations
-    Units are the MMTK standard, kJ/mol
+    """Uses APBS to calculate the solvation energy of a set of configurations.
+    Units are the MMTK standard, kJ/mol.
     """
     # Prepare configurations for writing to crd file
     factor = 1.0 / MMTK.Units.Ang

AlGDock/topology.py

@@ -46,7 +46,7 @@ def __init__(self, args, includeReceptor=False):
       os.path.dirname(args.FNs['ligand_database'])
     self.molecule = MMTK.Molecule(\
       os.path.basename(args.FNs['ligand_database']))
-    if includeReceptor:
+    if includeReceptor and (args.FNs['receptor_database']is not None):
       self.molecule_R = MMTK.Molecule(\
         os.path.basename(args.FNs['receptor_database']))
     else:
@@ -77,7 +77,10 @@ def __init__(self, args, includeReceptor=False):
     self.universe = MMTK.Universe.InfiniteUniverse()
     self.universe.addObject(self.molecule)
     if includeReceptor:
-      self.universe.addObject(self.molecule_R)
+      if self.molecule_R is not None:
+        self.universe.addObject(self.molecule_R)
+      else:
+        self.universe = None
 
     # Define L_first_atom
     if includeReceptor: