def main(list_input_dirs, energies_names, output_pre, scales, offsets):
#read in and rename arguments
inp_dir1=list_input_dirs[0][0]
scoretype1=list_input_dirs[0][1]
inp_dir2=list_input_dirs[1][0]
scoretype2=list_input_dirs[1][1]
title1 = os.path.basename(inp_dir1)
title2 = os.path.basename(inp_dir2)
#keep inp_dir1 as constant
#loop through energies_names[scoretype2]
#for each energies_names - make a list of weights either using random or centered around mean using sampling bins and nbins from input parameters
#recursive method - input is list of energies, list of weights per energy, and index of energy.
#for that energy, loops thru list of weights, and kicks off further method per next index
#base case - index = len(list). create weighted dec2, nat2. merge dicts. convert_disc and get list of discs, one per pdb. return this.
dec1_total = scorefileparse.read_dir(inp_dir1, scoretype1)
dec1_total_disc = discparse.pdbs_dict_to_metrics(dec1_total)
dec2_total = scorefileparse.read_dir(inp_dir2, scoretype2)
dec2_total_disc = discparse.pdbs_dict_to_metrics(dec2_total)
dec1 = scorefileparse.read_dir(inp_dir1, scoretype1, list_energies=energies_names[scoretype1])
weights = [ np.linspace(-2.0, 2.0, num=9).tolist() for _ in energies_names[scoretype2] ]
lwd = [ [] for _ in xrange(0, len(weights)+1) ] # need list for each weight + one for pdb_disc_dict
loop_weights(energies_names[scoretype2], weights, scales, offsets, dec1, inp_dir2, scoretype2, [], lwd)
list_avg_pdbs_deltaS_R = [ avg_pdbs_deltaS(deltaS_pdbs_disc_dict(pdbs_disc_dict, dec1_total_disc)) for pdbs_disc_dict in lwd[-1] ]
list_avg_pdbs_deltaS_A = [ avg_pdbs_deltaS(deltaS_pdbs_disc_dict(pdbs_disc_dict, dec2_total_disc)) for pdbs_disc_dict in lwd[-1] ]
list_weights_deltaS = lwd[:-1]
list_weights_deltaS.append(list_avg_pdbs_deltaS_R)
list_weights_deltaS.append(list_avg_pdbs_deltaS_A)
weights_deltaS_arr = np.array(list_weights_deltaS, dtype="float")
weights_deltaS_arr = weights_deltaS_arr.T
unique_z_vals = np.unique(weights_deltaS_arr[:,2])
low_dS = sorted(weights_deltaS_arr[:,3].tolist())[0:3]
low_dS_ind = [ i for i,dS in enumerate(weights_deltaS_arr[:,3].tolist()) if dS in low_dS ]
for ind in low_dS_ind:
pdvd.plot(dec1_total_disc, lwd[-1][ind], "Rosetta", "Combined", "{0}{1}_".format(output_pre,ind,"_"), add_slash=False)
pdvd.plot(dec2_total_disc, lwd[-1][ind], "Amber", "Combined", "{0}{1}_".format(output_pre,ind,"_"), add_slash=False)
#w_dS_sorted = weights_deltaS_arr[np.lexsort((weights_deltaS_arr[:,0],weights_deltaS_arr[:,1]))]
w_dS_sorted = weights_deltaS_arr
plot_weights_dS(w_dS_sorted, 3, unique_z_vals, title1, title2, energies_names[scoretype2][0], energies_names[scoretype2][1], energies_names[scoretype2][2])
plot_weights_dS(w_dS_sorted, 4, unique_z_vals, title1, title2, energies_names[scoretype2][0], energies_names[scoretype2][1], energies_names[scoretype2][2])
filename_txt = args.output_pre + "/" + title1 + "_" + title2 + "_weights_vs_deltaS.txt"
np.savetxt(filename_txt, w_dS_sorted, fmt='%2.5f', delimiter=",", header="gb,elec,dihedral,deltaSR,deltaSA")
def main(input_dir, output_pre, repl_orig):
#read in and rename arguments
title1 = os.path.basename(input_dir[0][0])
title2 = os.path.basename(input_dir[1][0])
d1 = scorefileparse.read_dir(input_dir[0][0], input_dir[0][1], repl_orig)
d2 = scorefileparse.read_dir(input_dir[1][0], input_dir[1][1], repl_orig)
[dec_inter1, dec_inter2] = scorefileparse.pdbs_scores_intersect([d1, d2])
disc_metrics_1 = discparse.pdbs_dict_to_metrics(dec_inter1,input_dir[0][1])
disc_metrics_2 = discparse.pdbs_dict_to_metrics(dec_inter2,input_dir[1][1])
print len(dec_inter1)
print len(dec_inter2)
print disc_metrics_2
plot(disc_metrics_1, disc_metrics_2, title1, title2, output_pre)
def main(input_dir, output_file):
#read in and rename arguments
title1 = os.path.basename(input_dir[0][0])
d1 = scorefileparse.read_dir(input_dir[0][0], input_dir[0][1], repl_orig=False)
disc_metrics = discparse.pdbs_dict_to_metrics(d1,input_dir[0][1])
discparse.show_pdbs_dict_metrics(disc_metrics, filename=output_file)
def main(input_dir_rosetta_sf, input_dir_amber_sf, input_dir_rosetta_pdb, input_dir_amber_pdb, output_dir, n_results):
d1 = scorefileparse.read_dir(input_dir_rosetta_sf, 'rosetta', repl_orig=True)
d2 = scorefileparse.read_dir(input_dir_amber_sf, 'amber', repl_orig=True)
dec_norm1 = scorefileparse.norm_pdbs(d1)
dec_norm2 = scorefileparse.norm_pdbs(d2)
[dec_inter1, dec_inter2] = scorefileparse.pdbs_scores_intersect([dec_norm1, dec_norm2])
dec1_cp = copy.deepcopy(dec_inter1)
dec2_cp = copy.deepcopy(dec_inter2)
min_naive_by_pdb = {}
for pdb in sorted(dec_inter1.keys()):
for i in range(1,n_results+1):
rosetta_lowest_energy = find_lowest_energy( dec_inter1[pdb] )
amber_lowest_energy = find_lowest_energy( dec_inter2[pdb] )
pareto_lowest_energy = find_pareto(dec1_cp, dec2_cp, pdb)
#copy rosetta file
src = os.join(input_dir_rosetta_pdb, pdb, rosetta_lowest_energy + "_0001.pdb")
dst = os.join(output_dir,"rosetta","{0}_{1}.pdb".format(pdb,i))
copyfile(src, dst)
#copy amber file
src = os.join(input_dir_rosetta_pdb, pdb, "min_NoH_" + amber_lowest_energy + ".pdb")
dst = os.join(output_dir,"amber","{0}_{1}.pdb".format(pdb,i))
copyfile(src, dst)
#copy pareto file
src = os.join(input_dir_rosetta_pdb, pdb, pareto_lowest_energy + "_0001.pdb")
dst = os.join(output_dir,"combined","{0}_{1}.pdb".format(pdb,i))
copyfile(src, dst)
#delete from original scoredict so that next round will get the next-lowest
dec_inter1[pdb].pop(rosetta_lowest_energy)
dec_inter2[pdb].pop(amber_lowest_energy)
dec1_cp[pdb].pop(pareto_lowest_energy)
dec2_cp[pdb].pop(pareto_lowest_energy)
def loop_weights(energies, weights, scales, offsets, dec1, inp_dir2, scoretype2, curr_weights, list_weight_discs):
if len(curr_weights) == len(energies):
dec2 = scorefileparse.read_dir(inp_dir2, scoretype2, list_energies=energies, weights=curr_weights, scales=scales, offsets=offsets)
[dec_inter1, dec_inter2] = scorefileparse.pdbs_intersect([dec1, dec2])
merged = scorefileparse.merge_pdbs_dicts([dec_inter1, dec_inter2])
list_discs = discparse.pdbs_dict_to_metrics(merged)
for ind, weight in enumerate(curr_weights):
list_weight_discs[ind].append(weight)
list_weight_discs[-1].append(list_discs)
return
energies_ind = len(curr_weights)
list_weights = weights[energies_ind]
for w in list_weights:
temp_weights = curr_weights[:]
temp_weights.append(w)
loop_weights(energies, weights, scales, offsets, dec1, inp_dir2, scoretype2, temp_weights, list_weight_discs)
def main(input_dir_rosetta_sf, input_dir_amber_sf, input_dir_rosetta_pdb, input_dir_amber_pdb, output_dir, n_results, alternate_rmsd, start_ind_desc_amber, start_ind_desc_rosetta, end_ind_desc_amber, end_ind_desc_rosetta, add_pdb_to_path):
if alternate_rmsd is True:
d1 = scorefileparse.read_dir(input_dir_rosetta_sf, 'rosetta', repl_orig=False, rmsd='total_score', start_ind_desc=start_ind_desc_rosetta, end_ind_desc=end_ind_desc_rosetta)
d2 = scorefileparse.read_dir(input_dir_amber_sf, 'amber', repl_orig=False, rmsd='tot', start_ind_desc=start_ind_desc_amber, end_ind_desc=end_ind_desc_amber)
else:
d1 = scorefileparse.read_dir(input_dir_rosetta_sf, 'rosetta', repl_orig=False, start_ind_desc=start_ind_desc_rosetta, end_ind_desc=end_ind_desc_rosetta)
d2 = scorefileparse.read_dir(input_dir_amber_sf, 'amber', repl_orig=False, start_ind_desc=start_ind_desc_amber, end_ind_desc=end_ind_desc_amber)
dec_norm1 = scorefileparse.norm_pdbs(d1)
dec_norm2 = scorefileparse.norm_pdbs(d2)
[dec_inter1, dec_inter2] = scorefileparse.pdbs_scores_intersect([dec_norm1, dec_norm2])
dec1_cp = copy.deepcopy(dec_inter1)
dec2_cp = copy.deepcopy(dec_inter2)
try:
os.mkdir( os.path.join(output_dir,"rosetta") )
except OSError:
pass
try:
os.mkdir( os.path.join(output_dir,"amber") )
except OSError:
pass
try:
os.mkdir( os.path.join(output_dir,"combined") )
except OSError:
pass
for pdb in sorted(dec_inter1.keys()):
for i in range(1,n_results+1):
rosetta_lowest_energy = find_lowest_energy( dec_inter1[pdb] )
amber_lowest_energy = find_lowest_energy( dec_inter2[pdb] )
pareto_lowest_energy = find_pareto(dec1_cp, dec2_cp, pdb)
if add_pdb_to_path == "True":
path_p = pdb
else:
path_p = ""
if end_ind_desc_rosetta == 1000:
r_suffix = ".pdb"
else:
r_suffix = "_0001.pdb"
if start_ind_desc_amber is None:
a_prefix = "min_NoH_"
else:
a_prefix = "min_"
if end_ind_desc_amber is None:
a_suffix = ".pdb"
else:
a_suffix = ".pdb.pdb"
#copy rosetta file
src = os.path.join(input_dir_rosetta_pdb, path_p, rosetta_lowest_energy + r_suffix)
dst = os.path.join(output_dir,"rosetta","{0}_{1}.pdb".format(pdb,i))
copyfile(src, dst)
#copy amber file
src = os.path.join(input_dir_amber_pdb, path_p, a_prefix + amber_lowest_energy + a_suffix)
dst = os.path.join(output_dir,"amber","{0}_{1}.pdb".format(pdb,i))
copyfile(src, dst)
#copy pareto file
src = os.path.join(input_dir_rosetta_pdb, path_p, pareto_lowest_energy + r_suffix)
dst = os.path.join(output_dir,"combined","{0}_{1}.pdb".format(pdb,i))
copyfile(src, dst)
#delete from original scoredict so that next round will get the next-lowest
dec_inter1[pdb].pop(rosetta_lowest_energy)
dec_inter2[pdb].pop(amber_lowest_energy)
dec1_cp[pdb].pop(pareto_lowest_energy)
dec2_cp[pdb].pop(pareto_lowest_energy)
def main(input_dir_rosetta_sf, input_dir_amber_sf, input_dir_rosetta_pdb,
input_dir_amber_pdb, output_dir, n_results, alternate_rmsd,
start_ind_desc_amber, start_ind_desc_rosetta, end_ind_desc_amber,
end_ind_desc_rosetta, add_pdb_to_path):
if alternate_rmsd is True:
d1 = scorefileparse.read_dir(input_dir_rosetta_sf,
'rosetta',
repl_orig=False,
rmsd='total_score',
start_ind_desc=start_ind_desc_rosetta,
end_ind_desc=end_ind_desc_rosetta)
d2 = scorefileparse.read_dir(input_dir_amber_sf,
'amber',
repl_orig=False,
rmsd='tot',
start_ind_desc=start_ind_desc_amber,
end_ind_desc=end_ind_desc_amber)
else:
d1 = scorefileparse.read_dir(input_dir_rosetta_sf,
'rosetta',
repl_orig=False,
start_ind_desc=start_ind_desc_rosetta,
end_ind_desc=end_ind_desc_rosetta)
d2 = scorefileparse.read_dir(input_dir_amber_sf,
'amber',
repl_orig=False,
start_ind_desc=start_ind_desc_amber,
end_ind_desc=end_ind_desc_amber)
dec_norm1 = scorefileparse.norm_pdbs(d1)
dec_norm2 = scorefileparse.norm_pdbs(d2)
[dec_inter1,
dec_inter2] = scorefileparse.pdbs_scores_intersect([dec_norm1, dec_norm2])
dec1_cp = copy.deepcopy(dec_inter1)
dec2_cp = copy.deepcopy(dec_inter2)
try:
os.mkdir(os.path.join(output_dir, "rosetta"))
except OSError:
pass
try:
os.mkdir(os.path.join(output_dir, "amber"))
except OSError:
pass
try:
os.mkdir(os.path.join(output_dir, "combined"))
except OSError:
pass
for pdb in sorted(dec_inter1.keys()):
for i in range(1, n_results + 1):
rosetta_lowest_energy = find_lowest_energy(dec_inter1[pdb])
amber_lowest_energy = find_lowest_energy(dec_inter2[pdb])
pareto_lowest_energy = find_pareto(dec1_cp, dec2_cp, pdb)
if add_pdb_to_path == "True":
path_p = pdb
else:
path_p = ""
if end_ind_desc_rosetta == 1000:
r_suffix = ".pdb"
else:
r_suffix = "_0001.pdb"
if start_ind_desc_amber is None:
a_prefix = "min_NoH_"
else:
a_prefix = "min_"
if end_ind_desc_amber is None:
a_suffix = ".pdb"
else:
a_suffix = ".pdb.pdb"
#copy rosetta file
src = os.path.join(input_dir_rosetta_pdb, path_p,
rosetta_lowest_energy + r_suffix)
dst = os.path.join(output_dir, "rosetta",
"{0}_{1}.pdb".format(pdb, i))
copyfile(src, dst)
#copy amber file
src = os.path.join(input_dir_amber_pdb, path_p,
a_prefix + amber_lowest_energy + a_suffix)
dst = os.path.join(output_dir, "amber",
"{0}_{1}.pdb".format(pdb, i))
copyfile(src, dst)
#copy pareto file
src = os.path.join(input_dir_rosetta_pdb, path_p,
pareto_lowest_energy + r_suffix)
dst = os.path.join(output_dir, "combined",
"{0}_{1}.pdb".format(pdb, i))
copyfile(src, dst)
#delete from original scoredict so that next round will get the next-lowest
dec_inter1[pdb].pop(rosetta_lowest_energy)
dec_inter2[pdb].pop(amber_lowest_energy)
dec1_cp[pdb].pop(pareto_lowest_energy)
dec2_cp[pdb].pop(pareto_lowest_energy)
