def gene_free_ene_2D(data_dict, num_sims, dim, Fx_old, rc_diml1, coefl1,
rc_diml2, coefl2, num_bins1, num_bins2, figname, xlabel1,
xlabel2, string_seq, colmap):
global KbT
data_RC1, bins_RC1, xaxis_RC1 = get_rc_data_1D(data_dict, num_sims,
rc_diml1, coefl1, num_bins1)
data_RC2, bins_RC2, xaxis_RC2 = get_rc_data_1D(data_dict, num_sims,
rc_diml2, coefl2, num_bins2)
# Generate the free energy for each bin
Prob_RC12 = [[0.0 for i in xrange(num_bins2)] for j in xrange(num_bins1)]
count = 0
for i in xrange(0, num_sims):
data_per_sim = len(data_dict[i + 1].data)
for j in xrange(0, data_per_sim):
# For the first dimension
each_data_RC1 = data_RC1[count]
each_count_RC1, bins_RCp1 = histogram(each_data_RC1, bins=bins_RC1)
count_indx1 = list(each_count_RC1).index(1)
# For the second dimension
each_data_RC2 = data_RC2[count]
each_count_RC2, bins_RCp2 = histogram(each_data_RC2, bins=bins_RC2)
count_indx2 = list(each_count_RC2).index(1)
#Unbias the free energy
Ubias = [0.0 for zero_val in xrange(num_sims)]
for k in xrange(num_sims):
for ii in xrange(0, dim):
equ_dis = data_dict[k + 1].equ_dis[ii]
constr = data_dict[k + 1].constr[ii]
samp_dis = data_dict[i + 1].data[j, ii]
Ubias[k] = Ubias[k] + 0.5 * constr * (samp_dis -
equ_dis)**2
denom = 0.0
for l in xrange(num_sims):
data_per_sim2 = len(data_dict[l + 1].data)
denom = denom + data_per_sim2 * exp(
(Fx_old[l] - Ubias[l]) / KbT)
Prob_RC12[count_indx1][count_indx2] = Prob_RC12[count_indx1][
count_indx2] + 1.0 / denom
count = count + 1
ini_crd = get_string_2D(data_dict, string_seq[0], string_seq[1], rc_diml1,
coefl1, rc_diml2, coefl2, figname + '.inistr')
fin_crd = get_string_2D(data_dict, string_seq[2], string_seq[3], rc_diml1,
coefl1, rc_diml2, coefl2, figname + '.laststr')
plot_free_ene_2D(num_bins1, num_bins2, xaxis_RC1, xaxis_RC2, Prob_RC12,
figname, xlabel1, xlabel2, colmap, ini_crd, fin_crd)
def gene_free_ene_2D(data_dict, num_sims, dim, expFx, Ubiasl, rc_diml1, coefl1,
rc_diml2, coefl2, num_bins1, num_bins2, figname, xlabel1,
xlabel2, string_seq, colmap):
data_RC1, bins_RC1, xaxis_RC1 = get_rc_data_1D(data_dict, num_sims,
rc_diml1, coefl1, num_bins1)
data_RC2, bins_RC2, xaxis_RC2 = get_rc_data_1D(data_dict, num_sims,
rc_diml2, coefl2, num_bins2)
# Generate the free energy for each bin
Prob_RC12 = [[0.0 for i in xrange(num_bins2)] for j in xrange(num_bins1)]
count = 0
kk = 0
for i in xrange(0, num_sims):
data_per_sim = len(data_dict[i + 1].data)
for l in xrange(0, data_per_sim):
# For the first dimension
each_data_RC1 = data_RC1[count]
each_count_RC1, bins_RCp1 = histogram(each_data_RC1, bins=bins_RC1)
count_indx1 = list(each_count_RC1).index(1)
# For the second dimension
each_data_RC2 = data_RC2[count]
each_count_RC2, bins_RCp2 = histogram(each_data_RC2, bins=bins_RC2)
count_indx2 = list(each_count_RC2).index(1)
#Get the biased potentials
Ubiasl_j = []
for j in xrange(num_sims):
Ubiasl_j.append(Ubiasl[kk])
kk = kk + 1
denom = 0.0
for k in xrange(num_sims):
data_per_sim2 = len(data_dict[k + 1].data)
denom = denom + float(data_per_sim2) * Ubiasl_j[k] * expFx[k]
Prob_RC12[count_indx1][count_indx2] = Prob_RC12[count_indx1][
count_indx2] + 1.0 / denom
count = count + 1
ini_crd = get_string_2D(data_dict, string_seq[0], string_seq[1], rc_diml1,
coefl1, rc_diml2, coefl2, figname + '.inistr')
fin_crd = get_string_2D(data_dict, string_seq[2], string_seq[3], rc_diml1,
coefl1, rc_diml2, coefl2, figname + '.laststr')
plot_free_ene_2D(num_bins1, num_bins2, xaxis_RC1, xaxis_RC2, Prob_RC12,
figname, xlabel1, xlabel2, colmap, ini_crd, fin_crd)
def gene_free_ene_1D(data_dict, num_sims, dim, Fx_old, rc_diml, coefl,
num_bins, figname, rc_label):
global KbT
data_RC, bins_RC, xaxis_RC = get_rc_data_1D(data_dict, num_sims, rc_diml,
coefl, num_bins)
# Generate the free energy for each bin
Prob_RC = [0.0 for i in xrange(num_bins)]
count = 0
for i in xrange(0, num_sims):
data_per_sim = len(data_dict[i + 1].data)
for j in xrange(0, data_per_sim):
each_data_RC = data_RC[count]
each_count_RC, bins_RCp = histogram(each_data_RC, bins=bins_RC)
count_indx = list(each_count_RC).index(1)
#Unbias the free energy
Ubias = [0.0 for zero_val in xrange(num_sims)]
for k in xrange(num_sims):
for ii in xrange(0, dim):
equ_dis = data_dict[k + 1].equ_dis[ii]
constr = data_dict[k + 1].constr[ii]
samp_dis = data_dict[i + 1].data[j, ii]
Ubias[k] = Ubias[k] + 0.5 * constr * (samp_dis -
equ_dis)**2
denom = 0.0
for l in xrange(num_sims):
data_per_sim2 = len(data_dict[l + 1].data)
denom = denom + data_per_sim2 * exp(
(Fx_old[l] - Ubias[l]) / KbT)
Prob_RC[count_indx] = Prob_RC[count_indx] + 1.0 / denom
count = count + 1
plot_free_ene_1D(num_bins, Prob_RC, xaxis_RC, figname, rc_label)
def gene_free_ene_1D(data_dict, num_sims, dim, expFx, Ubiasl, rc_diml, coefl,
num_bins, figname, rc_label):
"""The part is from eqs 10 and 11 in the WHAM paper"""
data_RC, bins_RC, xaxis_RC = get_rc_data_1D(data_dict, num_sims, rc_diml,
coefl, num_bins)
# Generate the free energy for each bin
Prob_RC = [0.0 for i in xrange(num_bins)]
count = 0
kk = 0
for i in xrange(0, num_sims): #Sum over big N for i
data_per_sim = len(data_dict[i + 1].data)
for l in xrange(0, data_per_sim): #Sum over little n for l
#Calculate the probability for each point
each_data_RC = data_RC[count]
each_count_RC, bins_RCp = histogram(each_data_RC, bins=bins_RC)
count_indx = list(each_count_RC).index(1)
#Get the biased potentials
Ubiasl_j = []
for j in xrange(num_sims):
Ubiasl_j.append(Ubiasl[kk])
kk = kk + 1
#Obtain the denominator
denom = 0.0
for k in xrange(num_sims):
data_per_sim2 = len(data_dict[k + 1].data)
denom = denom + float(data_per_sim2) * Ubiasl_j[k] * expFx[k]
Prob_RC[count_indx] = Prob_RC[count_indx] + 1.0 / denom
count = count + 1
plot_free_ene_1D(num_bins, Prob_RC, xaxis_RC, figname, rc_label)
def gene_free_ene_2D(data_dict, num_sims, dim, expFx, Ubiasl, rc_diml1, coefl1,
rc_diml2, coefl2, num_bins1, num_bins2, figname, xlabel1, xlabel2,
string_seq, colmap, avg=0, tot_ene=[], ene1=[]):
data_RC1, bins_RC1, xaxis_RC1 = get_rc_data_1D(data_dict, num_sims, rc_diml1, coefl1, num_bins1)
data_RC2, bins_RC2, xaxis_RC2 = get_rc_data_1D(data_dict, num_sims, rc_diml2, coefl2, num_bins2)
if avg == 0: # No average to calculate
pass
elif avg == 1: # Only average to calculate
if tot_ene == []:
raise ValueError('No value list provided to calculate the average value!')
elif tot_ene != []:
data_list = tot_ene
if ene1 != []:
raise ValueError('Provide two list values to calculate the average '
'value, only one list is needed!')
elif avg == 2: # About the free energy partitioning
if tot_ene != [] and ene1 != []:
if len(tot_ene) == len(ene1):
#Calculate the exp[beta*(U(ri)-U'(ri))]
ene2 = [tot_ene[i]-ene1[i] for i in xrange(0, len(tot_ene))]
ene2min = min(ene2)
ene2 = [i-ene2min-300.0 for i in ene2]
data_list = [exp(i/KbT) for i in ene2]
else:
raise ValueError('The length of total energy list and paritioning '
'energy list are not the same!')
elif tot_ene != []:
raise ValueError('Only total energy list provided, you need to '
'provide paritioning energy list as well!')
elif ene1 != []:
raise ValueError('Only paritioning energy list provided, you need to '
'provide total energy list as well!')
# Generate the free energy for each bin
Prob_RC12_0 = [[0.0 for i in xrange(num_bins2)] for j in xrange(num_bins1)]
Prob_RC12_1 = [[0.0 for i in xrange(num_bins2)] for j in xrange(num_bins1)]
count = 0
kk = 0
for i in xrange(0, num_sims):
data_per_sim = len(data_dict[i+1].data)
for l in xrange(0, data_per_sim):
# For the first dimension
each_data_RC1 = data_RC1[count]
each_count_RC1, bins_RCp1 = histogram(each_data_RC1, bins=bins_RC1)
count_indx1 = list(each_count_RC1).index(1)
# For the second dimension
each_data_RC2 = data_RC2[count]
each_count_RC2, bins_RCp2 = histogram(each_data_RC2, bins=bins_RC2)
count_indx2 = list(each_count_RC2).index(1)
#Get the biased potentials
Ubiasl_j = []
for j in xrange(num_sims):
Ubiasl_j.append(Ubiasl[kk])
kk = kk + 1
denom = 0.0
for k in xrange(num_sims):
data_per_sim2 = len(data_dict[k+1].data)
denom = denom + float(data_per_sim2) * Ubiasl_j[k] * expFx[k]
Prob_RC12_0[count_indx1][count_indx2] = Prob_RC12_0[count_indx1][count_indx2] + 1.0/denom
Prob_RC12_1[count_indx1][count_indx2] = Prob_RC12_1[count_indx1][count_indx2] + data_list[count]/denom
count = count + 1
ini_crd = get_string_2D(data_dict, string_seq[0], string_seq[1], rc_diml1, coefl1, rc_diml2, coefl2, figname + '.inistr')
fin_crd = get_string_2D(data_dict, string_seq[2], string_seq[3], rc_diml1, coefl1, rc_diml2, coefl2, figname + '.laststr')
if avg==0:
plot_free_ene_2D(num_bins1, num_bins2, xaxis_RC1, xaxis_RC2, Prob_RC12_0, figname, xlabel1, xlabel2, colmap, ini_crd, fin_crd, 0)
else:
Prob_RC12_2 = [[0.0 for i in xrange(num_bins2)] for j in xrange(num_bins1)]
for i in xrange(num_bins2):
for j in xrange(num_bins1):
if Prob_RC12_0[j][i] != 0:
Prob_RC12_2[j][i] = Prob_RC12_1[j][i]/Prob_RC12_0[j][i]
plot_free_ene_2D(num_bins1, num_bins2, xaxis_RC1, xaxis_RC2, Prob_RC12_2, figname, xlabel1, xlabel2, colmap, ini_crd, fin_crd, avg)
def gene_free_ene_1D(data_dict, num_sims, dim, expFx, Ubiasl, rc_diml, coefl, num_bins, figname, rc_label, avg=0, tot_ene=[], ene1=[]):
"""The part is from eqs 10 and 11 in the WHAM paper"""
data_RC, bins_RC, xaxis_RC = get_rc_data_1D(data_dict, num_sims, rc_diml, coefl, num_bins)
if avg == 0: # No average to calculate
pass
elif avg == 1: # Only average to calculate
if tot_ene == []:
raise ValueError('No value list provided to calculate the average value!')
elif tot_ene != []:
data_list = tot_ene
if ene1 != []:
raise ValueError('Provide two list values to calculate the average '
'value, only one list is needed!')
elif avg == 2: # About the free energy partitioning
if tot_ene != [] and ene1 != []:
if len(tot_ene) == len(ene1):
#Calculate the exp[beta*(U(ri)-U'(ri))]
ene2 = [tot_ene[i]-ene1[i] for i in xrange(0, len(tot_ene))]
ene2min = min(ene2)
ene2 = [i-ene2min-300.0 for i in ene2]
data_list = [exp(i/KbT) for i in ene2]
else:
raise ValueError('The length of total energy list and paritioning '
'energy list are not the same!')
elif tot_ene != []:
raise ValueError('Only total energy list provided, you need to '
'provide paritioning energy list as well!')
elif ene1 != []:
raise ValueError('Only paritioning energy list provided, you need to '
'provide total energy list as well!')
# Generate the free energy for each bin
Prob_RC0 = [0.0 for i in xrange(num_bins)]
Prob_RC1 = [0.0 for i in xrange(num_bins)]
count = 0
kk = 0
for i in xrange(0, num_sims): #Sum over big N for i
data_per_sim = len(data_dict[i+1].data)
for l in xrange(0, data_per_sim): #Sum over little n for l
#Calculate the probability for each point
each_data_RC = data_RC[count]
each_count_RC, bins_RCp = histogram(each_data_RC, bins=bins_RC)
count_indx = list(each_count_RC).index(1)
#Get the biased potentials
Ubiasl_j = []
for j in xrange(num_sims):
Ubiasl_j.append(Ubiasl[kk])
kk = kk + 1
#Obtain the denominator
denom = 0.0
for k in xrange(num_sims):
data_per_sim2 = len(data_dict[k+1].data)
denom = denom + float(data_per_sim2) * Ubiasl_j[k] * expFx[k]
Prob_RC0[count_indx] = Prob_RC0[count_indx] + 1.0/denom
Prob_RC1[count_indx] = Prob_RC1[count_indx] + data_list[count]/denom
count = count + 1
if avg==0:
plot_free_ene_1D(num_bins, Prob_RC0, xaxis_RC, figname, rc_label, 0)
else:
Prob_RC2 = [Prob_RC1[i]/Prob_RC0[i] for i in xrange(num_bins)]
plot_free_ene_1D(num_bins, Prob_RC2, xaxis_RC, figname, rc_label, avg)