Пример #1
0
def main():

	usage_txt = "\nTakes trr-trajectory and internals-definition and outputs timeseries as npz- or mat-file.\nOutput-filetype is decided based on given filename-extension.\n\n   %prog [options] <trr-input-file> <int-file> <output-file>"
	parser = OptionParser(usage=usage_txt, version="%prog 0.1")
	parser.add_option("-d", "--degrees", action="store_true", default=False, help="save dihedrals in [deg]")

	(options, args) = parser.parse_args()

	if(len(args) != 3):
		parser.error("incorrect number of arguments")

	if(options.degrees):
		print("\nDihedral angles will be saved in [deg].\n")
	else:
		print("\nDihedral angles will be saved in [rad].\n")

	trr_fn = args[0]
	internals_fn = args[1]
	out_fn = args[2]

	converter = Converter(internals_fn)
	all_frames_int = converter.read_trajectory(trr_fn)
	print("loaded trr-file with %d frames" % len(all_frames_int))

	if(options.degrees):
		all_frames_int.dihedral_array = np.degrees(all_frames_int.dihedral_array)

	dih_labels = [c.label for c in converter.dihedrals]
	lin_labels = [c.label for c in converter.linears]

	if(out_fn.endswith(".npz")):
		print("Writing NumPy file: "+out_fn)
		if (all_frames_int.has_dihedrals and all_frames_int.has_linears):
			np.savez(out_fn, linears=all_frames_int.linear_array, linear_labels=lin_labels, dihedrals=all_frames_int.dihedral_array, dihedral_labels=dih_labels)
		elif(all_frames_int.has_dihedrals):
			np.savez(out_fn, dihedrals=all_frames_int.dihedral_array, dihedral_labels=dih_labels)
		else:
			np.savez(out_fn, linears=all_frames_int.linear_array, linear_labels=lin_labels)	
	elif(out_fn.endswith(".mat")):
		print("Writing Matlab file: "+out_fn)
		if (all_frames_int.has_dihedrals and all_frames_int.has_linears):
			savemat(out_fn, {'linears':all_frames_int.linear_array, 'dihedrals':all_frames_int.dihedral_array})
		elif(all_frames_int.has_dihedrals):
			savemat(out_fn, {'dihedrals':all_frames_int.dihedral_array})
		else:
			savemat(out_fn, {'linears':all_frames_int.linear_array})
	else:
		raise(Exception("Unkown output filetype: "+out_fn))
Пример #2
0
 def converter(self):  #TODO maybe cache instance
     if (hasattr(self, "int_fn")):
         return (Converter(self.int_fn))
     return (None)
Пример #3
0
def main():
    options = options_desc.parse_args(sys.argv)[0]

    if options.common_filename:
        options.molecule = options.common_filename + ".pdb"
        options.presampling = options.common_filename + ".trr"
        options.internals = options.common_filename + ".int"
        options.grompp = options.common_filename + ".mdp"
        options.topology = options.common_filename + ".top"
        options.index = options.common_filename + ".ndx"

    print("Options:\n%s\n" % pformat(eval(str(options))))

    assert path.exists(options.molecule)
    assert path.exists(options.presampling)
    assert path.exists(options.internals)
    assert path.exists(options.grompp)
    assert path.exists(options.topology)

    # TODO: what if there is no index-file? (make_ndx)
    assert path.exists(options.index)
    assert "MOI" in gromacs.read_index_file(options.index), "group MOI should be defined in index file"

    # checks e.g. if the mdp-file looks good
    mdp_options = gromacs.read_mdp_file(options.grompp)

    # options we cannot fix
    for ref_t in re.findall("[0-9]+", mdp_options["ref_t"]):
        assert int(ref_t) == options.temperature, "temperature in mdp file does not match ZIBgridfree temperature"
        # TODO drop options.temperature and get temperature directly from mdp file... ask again if temperature is above 310K

        # options we can fix
    mdp_options_dirty = False  # if set, a new mdp-file will be written
    required_mdp_options = {"dihre": "yes", "dihre_fc": "1", "disre": "simple", "disre_fc": "1"}
    for (k, v) in required_mdp_options.items():
        if mdp_options.has_key(k):
            assert mdp_options[k] == v  # check, if we would overwrite something
        else:
            mdp_options[k] = v
            mdp_options_dirty = True

    if mdp_options.has_key("energygrps"):
        assert "MOI" in [
            str(egrp) for egrp in re.findall("[\S]+", mdp_options["energygrps"])
        ], "group MOI should be among energygrps in mdp file"
    else:
        mdp_options["energygrps"] = "MOI"
        mdp_options_dirty = True

    a, b = mdp_options.has_key("nstxout"), mdp_options.has_key("nstenergy")
    if a and not b:
        mdp_options["nstenergy"] = mdp_options["nstxout"]
        mdp_options_dirty = True
    elif b and not a:
        mdp_options["nstxout"] = mdp_options["nstenergy"]
        mdp_options_dirty = True
    elif b and a:
        assert mdp_options["nstxout"] == mdp_options["nstenergy"], "nstxout should equal nstenergy"

    if int(mdp_options["nsteps"]) > 1e6:
        msg = "Number of MD-steps?"
        mdp_options["nsteps"] = str(userinput(msg, "int", default=int(mdp_options["nsteps"])))

        # create a fixed mdp-file
    if mdp_options_dirty:
        print("Creating copy of mdp-file and adding missing options.")
        out_fn = options.grompp.rsplit(".", 1)[0] + "_fixed.mdp"
        f = open(out_fn, "w")  # append
        f.write("; Generated by zgf_create_pool\n")
        for i in mdp_options.items():
            f.write("%s = %s\n" % i)
        f.write("; EOF\n")
        f.close()
        options.grompp = out_fn

        # check if subsampling is reasonable
    if os.path.getsize(options.presampling) > 100e6:  # 100MB
        print("Presampling trajectory is large")
        trr = TrrFile(options.presampling)
        dt = trr.first_frame.next().t - trr.first_frame.t
        trr.close()
        print("Presampling timestep is %.2f ps" % dt)
        if dt < 10:  # picoseconds
            # TODO: maybe calculate subsampling factor individually, or ask?
            msg = "Subsample presampling trajectory by a tenth?"
            if userinput(msg, "bool"):
                out_fn = options.presampling.rsplit(".", 1)[0] + "_tenth.trr"
                cmd = ["trjconv", "-f", options.presampling, "-o", out_fn, "-skip", "10"]
                check_call(cmd)
                options.presampling = out_fn

                # balance linears
    if options.balance_linears:
        print("Balance Linears")
        old_converter = Converter(options.internals)
        print("Loading presampling....")
        frames = old_converter.read_trajectory(options.presampling)
        new_coord_list = []
        for c in old_converter:
            if not isinstance(c, LinearCoordinate):
                new_coord_list.append(c)
                continue  # we do not work on other Coordinate-Types
                # TODO: is this a good way to determine new_weight and new_offset???
            new_weight = c.weight / sqrt(2 * frames.var().getcoord(c))
            new_offset = c.offset + frames.mean().getcoord(c)
            new_coord = LinearCoordinate(*c.atoms, label=c.label, weight=new_weight, offset=new_offset)
            new_coord_list.append(new_coord)
        new_converter = Converter(coord_list=new_coord_list)

        assert old_converter.filename.endswith(".int")
        options.internals = old_converter.filename[:-4] + "_balanced.int"
        print("Writing balanced Converter to: " + options.internals)
        f = open(options.internals, "w")
        f.write(new_converter.serialize())
        f.close()
        assert len(Converter(options.internals)) == len(new_coord_list)  # try parsing

        # Finally: Create root-node and pool
    pool = Pool()
    if len(pool) != 0:
        print("ERROR: A pool already exists here.")
        sys.exit(1)

    pool.int_fn = options.internals
    pool.mdp_fn = options.grompp
    pool.top_fn = options.topology
    pool.ndx_fn = options.index
    pool.temperature = options.temperature
    pool.gr_threshold = options.gr_threshold
    pool.gr_chains = options.gr_chains
    pool.alpha = None
    pool.save()  # save pool for the first time...

    # ... then we can save the first node...
    node0 = Node()
    node0.state = "refined"
    node0.save()  # also creates the node directory ... needed for symlink
    os.symlink(os.path.relpath(options.presampling, node0.dir), node0.trr_fn)
    os.symlink(os.path.relpath(options.molecule, node0.dir), node0.pdb_fn)

    pool.root_name = node0.name
    pool.save()  # ... now we have to save the pool again.

    if not path.exists("analysis"):
        os.mkdir("analysis")
Пример #4
0
def main():
	options = options_desc.parse_args(sys.argv)[0]

	if(options.common_filename):
		options.molecule = options.common_filename+".pdb"
		options.presampling = options.common_filename+".trr"
		options.internals = options.common_filename+".int"
		options.grompp = options.common_filename+".mdp"
		options.topology = options.common_filename+".top"
		options.index = options.common_filename+".ndx"

	print("Options:\n%s\n"%pformat(eval(str(options))))

	assert(path.exists(options.molecule))
	assert(path.exists(options.presampling))
	assert(path.exists(options.internals))
	assert(path.exists(options.grompp))
	assert(path.exists(options.topology))
		
	#TODO: what if there is no index-file? (make_ndx)
	assert(path.exists(options.index))
	assert('moi' in gromacs.read_index_file(options.index)), "group 'MOI' should be defined in index file"
 
	# checks e.g. if the mdp-file looks good
	mdp_options = gromacs.read_mdp_file(options.grompp)
	
	temperatures = [ref_t for ref_t in re.findall("[0-9]+", mdp_options["ref_t"])]
	assert(len(set(temperatures)) == 1), "temperature definition in mdp file is ambiguous"
	temperature = temperatures[0]

	# get sampling temperature from mdp file
	if(int(temperature) > 310):
		if not(userinput("Your sampling temperature is set to %s K. Continue?"%temperature, "bool")):
			sys.exit("Quit by user.")

	# options we can fix 
 	mdp_options_dirty = False #if set, a new mdp-file will be written

	# the value of the following options need to be fixed
	critical_mdp_options = {"dihre":"yes", "dihre_fc":"1", "disre":"simple", "disre_fc":"1", "gen_temp":temperature}
	for (k,v) in critical_mdp_options.items():
 		if(mdp_options.has_key(k) and mdp_options[k].strip() != v):
			print "Error. I do not want to use '%s' for option '%s' ('%s' required). Please fix your mdp file."%(mdp_options[k].strip(),k,v)
			sys.exit("Quitting.")
 		else:
 			mdp_options[k] = v
 			mdp_options_dirty = True

	# the value of the following options does not matter, but they should be there
	noncritical_mdp_options = {"tcoupl":"no", "pcoupl":"no", "gen_vel":"no", "gen_seed":"-1"}
	for (k,v) in noncritical_mdp_options.items():
		if not(mdp_options.has_key(k)):
			mdp_options[k] = v
			mdp_options_dirty = True

	a = mdp_options.has_key("energygrps") and "moi" not in [str(egrp) for egrp in re.findall('[\S]+', mdp_options["energygrps"])]
	b = not(mdp_options.has_key("energygrps"))
	if(a or b):
		if not(userinput("'MOI' is not defined as an energy group in your mdp file. Maybe you have forgotten to define proper 'energygrps'. Continue?", "bool")):
			sys.exit("Quit by user.")

	a, b = mdp_options.has_key("nstxout"), mdp_options.has_key("nstenergy")
	if(a and not b):
		mdp_options["nstenergy"] = mdp_options["nstxout"]
		mdp_options_dirty = True
	elif(b and not a):
		mdp_options["nstxout"] = mdp_options["nstenergy"]
		mdp_options_dirty = True
	elif(b and a):
		assert(mdp_options["nstxout"] == mdp_options["nstenergy"]), "nstxout should equal nstenergy"
		
	if(int(mdp_options["nsteps"]) > 1e6):
		msg = "Number of MD-steps?"
		mdp_options["nsteps"] = str( userinput(msg, "int", default=int(mdp_options["nsteps"])) )
	
	# create a fixed mdp-file
	if(mdp_options_dirty):
		print("Creating copy of mdp-file and adding missing options.")
		out_fn = options.grompp.rsplit(".", 1)[0] + "_fixed.mdp"
		f = open(out_fn, "w") # append
		f.write("; Generated by zgf_create_pool\n")
		for i in sorted(mdp_options.items()):
			f.write("%s = %s\n"%i)
		f.write("; EOF\n")
		f.close()
		options.grompp = out_fn
		
	
	# check if subsampling is reasonable
	if(os.path.getsize(options.presampling) > 100e6): # 100MB
		print("Presampling trajectory is large")
		trr = TrrFile(options.presampling)
		dt = trr.first_frame.next().t - trr.first_frame.t
		trr.close()
		print("Presampling timestep is %.2f ps"%dt)
		if(dt < 10): # picoseconds
			#TODO: maybe calculate subsampling factor individually, or ask? 
			msg = "Subsample presampling trajectory by a tenth?"
			if(userinput(msg, "bool")):
				out_fn = options.presampling.rsplit(".", 1)[0] + "_tenth.trr"
				cmd = ["trjconv", "-f", options.presampling, "-o", out_fn, "-skip", "10"]
				check_call(cmd)
				options.presampling = out_fn
	
			
	# balance linears
	if(options.balance_linears):
		print("Balance Linears")
		old_converter = Converter(options.internals)
		print("Loading presampling....")
		frames = old_converter.read_trajectory(options.presampling)
		new_coord_list = []
		for c in old_converter:
			if(not isinstance(c, LinearCoordinate)):
				new_coord_list.append(c)
				continue # we do not work on other Coordinate-Types
			#TODO: is this a good way to determine new_weight and new_offset??? 
			new_weight = c.weight / sqrt(2*frames.var().getcoord(c))
			new_offset = c.offset + frames.mean().getcoord(c)
			new_coord = LinearCoordinate(*c.atoms, label=c.label, weight=new_weight, offset=new_offset)
			new_coord_list.append(new_coord)
		new_converter = Converter(coord_list=new_coord_list)
	
		assert(old_converter.filename.endswith(".int"))
		options.internals = old_converter.filename[:-4] + "_balanced.int"
		print("Writing balanced Converter to: "+options.internals)
		f = open(options.internals, "w")
		f.write(new_converter.serialize())
		f.close()
		assert(len(Converter(options.internals)) == len(new_coord_list)) #try parsing
	
	# Finally: Create root-node and pool
	pool = Pool()
	if(len(pool) != 0):
		print("ERROR: A pool already exists here.")
		sys.exit(1)
	
	pool.int_fn = options.internals
	pool.mdp_fn = options.grompp
	pool.top_fn = options.topology
	pool.ndx_fn = options.index
	pool.temperature = int(temperature)
	pool.gr_threshold = options.gr_threshold
	pool.gr_chains = options.gr_chains
	pool.alpha = None
	pool.save() # save pool for the first time...

	# ... then we can save the first node...
	node0 = Node()
	node0.state = "refined"	
	node0.save() # also creates the node directory ... needed for symlink
	os.symlink(os.path.relpath(options.presampling, node0.dir), node0.trr_fn)
	os.symlink(os.path.relpath(options.molecule, node0.dir), node0.pdb_fn)
	
	pool.root_name = node0.name
	pool.save() #... now we have to save the pool again.
	
	if(not path.exists("analysis")):
		os.mkdir("analysis")
Пример #5
0
def main():
    options = options_desc.parse_args(sys.argv)[0]

    if (options.common_filename):
        options.molecule = options.common_filename + ".pdb"
        options.presampling = options.common_filename + ".trr"
        options.internals = options.common_filename + ".int"
        options.grompp = options.common_filename + ".mdp"
        options.topology = options.common_filename + ".top"
        options.index = options.common_filename + ".ndx"

    print("Options:\n%s\n" % pformat(eval(str(options))))

    assert (path.exists(options.molecule))
    assert (path.exists(options.presampling))
    assert (path.exists(options.internals))
    assert (path.exists(options.grompp))
    assert (path.exists(options.topology))

    #TODO: what if there is no index-file? (make_ndx)
    assert (path.exists(options.index))
    assert ('moi' in gromacs.read_index_file(
        options.index)), "group 'MOI' should be defined in index file"

    # checks e.g. if the mdp-file looks good
    mdp_options = gromacs.read_mdp_file(options.grompp)

    temperatures = [
        ref_t for ref_t in re.findall("[0-9]+", mdp_options["ref_t"])
    ]
    assert (len(set(temperatures)) == 1
            ), "temperature definition in mdp file is ambiguous"
    temperature = temperatures[0]

    # get sampling temperature from mdp file
    if (int(temperature) > 310):
        if not (userinput(
                "Your sampling temperature is set to %s K. Continue?" %
                temperature, "bool")):
            sys.exit("Quit by user.")

    # options we can fix
    mdp_options_dirty = False  #if set, a new mdp-file will be written

    # the value of the following options need to be fixed
    critical_mdp_options = {
        "dihre": "yes",
        "dihre_fc": "1",
        "disre": "simple",
        "disre_fc": "1",
        "gen_temp": temperature
    }
    for (k, v) in critical_mdp_options.items():
        if (mdp_options.has_key(k) and mdp_options[k].strip() != v):
            print "Error. I do not want to use '%s' for option '%s' ('%s' required). Please fix your mdp file." % (
                mdp_options[k].strip(), k, v)
            sys.exit("Quitting.")
        else:
            mdp_options[k] = v
            mdp_options_dirty = True

    # the value of the following options does not matter, but they should be there
    noncritical_mdp_options = {
        "tcoupl": "no",
        "pcoupl": "no",
        "gen_vel": "no",
        "gen_seed": "-1"
    }
    for (k, v) in noncritical_mdp_options.items():
        if not (mdp_options.has_key(k)):
            mdp_options[k] = v
            mdp_options_dirty = True

    a = mdp_options.has_key("energygrps") and "moi" not in [
        str(egrp) for egrp in re.findall('[\S]+', mdp_options["energygrps"])
    ]
    b = not (mdp_options.has_key("energygrps"))
    if (a or b):
        if not (userinput(
                "'MOI' is not defined as an energy group in your mdp file. Maybe you have forgotten to define proper 'energygrps'. Continue?",
                "bool")):
            sys.exit("Quit by user.")

    a, b = mdp_options.has_key("nstxout"), mdp_options.has_key("nstenergy")
    if (a and not b):
        mdp_options["nstenergy"] = mdp_options["nstxout"]
        mdp_options_dirty = True
    elif (b and not a):
        mdp_options["nstxout"] = mdp_options["nstenergy"]
        mdp_options_dirty = True
    elif (b and a):
        assert (mdp_options["nstxout"] == mdp_options["nstenergy"]
                ), "nstxout should equal nstenergy"

    if (int(mdp_options["nsteps"]) > 1e6):
        msg = "Number of MD-steps?"
        mdp_options["nsteps"] = str(
            userinput(msg, "int", default=int(mdp_options["nsteps"])))

    # create a fixed mdp-file
    if (mdp_options_dirty):
        print("Creating copy of mdp-file and adding missing options.")
        out_fn = options.grompp.rsplit(".", 1)[0] + "_fixed.mdp"
        f = open(out_fn, "w")  # append
        f.write("; Generated by zgf_create_pool\n")
        for i in sorted(mdp_options.items()):
            f.write("%s = %s\n" % i)
        f.write("; EOF\n")
        f.close()
        options.grompp = out_fn

    # check if subsampling is reasonable
    if (os.path.getsize(options.presampling) > 100e6):  # 100MB
        print("Presampling trajectory is large")
        trr = TrrFile(options.presampling)
        dt = trr.first_frame.next().t - trr.first_frame.t
        trr.close()
        print("Presampling timestep is %.2f ps" % dt)
        if (dt < 10):  # picoseconds
            #TODO: maybe calculate subsampling factor individually, or ask?
            msg = "Subsample presampling trajectory by a tenth?"
            if (userinput(msg, "bool")):
                out_fn = options.presampling.rsplit(".", 1)[0] + "_tenth.trr"
                cmd = [
                    "trjconv", "-f", options.presampling, "-o", out_fn,
                    "-skip", "10"
                ]
                check_call(cmd)
                options.presampling = out_fn

    # balance linears
    if (options.balance_linears):
        print("Balance Linears")
        old_converter = Converter(options.internals)
        print("Loading presampling....")
        frames = old_converter.read_trajectory(options.presampling)
        new_coord_list = []
        for c in old_converter:
            if (not isinstance(c, LinearCoordinate)):
                new_coord_list.append(c)
                continue  # we do not work on other Coordinate-Types
            #TODO: is this a good way to determine new_weight and new_offset???
            new_weight = c.weight / sqrt(2 * frames.var().getcoord(c))
            new_offset = c.offset + frames.mean().getcoord(c)
            new_coord = LinearCoordinate(*c.atoms,
                                         label=c.label,
                                         weight=new_weight,
                                         offset=new_offset)
            new_coord_list.append(new_coord)
        new_converter = Converter(coord_list=new_coord_list)

        assert (old_converter.filename.endswith(".int"))
        options.internals = old_converter.filename[:-4] + "_balanced.int"
        print("Writing balanced Converter to: " + options.internals)
        f = open(options.internals, "w")
        f.write(new_converter.serialize())
        f.close()
        assert (len(Converter(options.internals)) == len(new_coord_list)
                )  #try parsing

    # Finally: Create root-node and pool
    pool = Pool()
    if (len(pool) != 0):
        print("ERROR: A pool already exists here.")
        sys.exit(1)

    pool.int_fn = options.internals
    pool.mdp_fn = options.grompp
    pool.top_fn = options.topology
    pool.ndx_fn = options.index
    pool.temperature = int(temperature)
    pool.gr_threshold = options.gr_threshold
    pool.gr_chains = options.gr_chains
    pool.alpha = None
    pool.save()  # save pool for the first time...

    # ... then we can save the first node...
    node0 = Node()
    node0.state = "refined"
    node0.save()  # also creates the node directory ... needed for symlink
    os.symlink(os.path.relpath(options.presampling, node0.dir), node0.trr_fn)
    os.symlink(os.path.relpath(options.molecule, node0.dir), node0.pdb_fn)

    pool.root_name = node0.name
    pool.save()  #... now we have to save the pool again.

    if (not path.exists("analysis")):
        os.mkdir("analysis")
Пример #6
0
def main():

    usage_txt = "\nTakes trr-trajectory and internals-definition and outputs timeseries as npz- or mat-file.\nOutput-filetype is decided based on given filename-extension.\n\n   %prog [options] <trr-input-file> <int-file> <output-file>"
    parser = OptionParser(usage=usage_txt, version="%prog 0.1")
    parser.add_option("-d",
                      "--degrees",
                      action="store_true",
                      default=False,
                      help="save dihedrals in [deg]")

    (options, args) = parser.parse_args()

    if (len(args) != 3):
        parser.error("incorrect number of arguments")

    if (options.degrees):
        print("\nDihedral angles will be saved in [deg].\n")
    else:
        print("\nDihedral angles will be saved in [rad].\n")

    trr_fn = args[0]
    internals_fn = args[1]
    out_fn = args[2]

    converter = Converter(internals_fn)
    all_frames_int = converter.read_trajectory(trr_fn)
    print("loaded trr-file with %d frames" % len(all_frames_int))

    if (options.degrees):
        all_frames_int.dihedral_array = np.degrees(
            all_frames_int.dihedral_array)

    dih_labels = [c.label for c in converter.dihedrals]
    lin_labels = [c.label for c in converter.linears]

    if (out_fn.endswith(".npz")):
        print("Writing NumPy file: " + out_fn)
        if (all_frames_int.has_dihedrals and all_frames_int.has_linears):
            np.savez(out_fn,
                     linears=all_frames_int.linear_array,
                     linear_labels=lin_labels,
                     dihedrals=all_frames_int.dihedral_array,
                     dihedral_labels=dih_labels)
        elif (all_frames_int.has_dihedrals):
            np.savez(out_fn,
                     dihedrals=all_frames_int.dihedral_array,
                     dihedral_labels=dih_labels)
        else:
            np.savez(out_fn,
                     linears=all_frames_int.linear_array,
                     linear_labels=lin_labels)
    elif (out_fn.endswith(".mat")):
        print("Writing Matlab file: " + out_fn)
        if (all_frames_int.has_dihedrals and all_frames_int.has_linears):
            savemat(
                out_fn, {
                    'linears': all_frames_int.linear_array,
                    'dihedrals': all_frames_int.dihedral_array
                })
        elif (all_frames_int.has_dihedrals):
            savemat(out_fn, {'dihedrals': all_frames_int.dihedral_array})
        else:
            savemat(out_fn, {'linears': all_frames_int.linear_array})
    else:
        raise (Exception("Unkown output filetype: " + out_fn))