def read_train_data():
pro_list = []
lig_list = []
for index in range(1, 2):
digit = format(index, '04d')
file_name_pro = 'training_data/' + digit + '_pro_cg.pdb'
file_name_lig = 'training_data/' + digit + '_lig_cg.pdb'
pro = read_pdb(file_name_pro)
lig = read_pdb(file_name_lig)
pro_list.append(pro)
lig_list.append(lig)
def read_raw_data(n, file_type):
if (n < 1) or (n > 3000):
return "Please provide a valid number of files between 1 to 3000"
if (file_type != "lig") and (file_type != "pro"):
return "Please provide a valid file type which should be 'lig' or 'pro'"
res = []
for i in range(1, n + 1):
file_number = ("0000" + str(i))[-4:]
file_name = "{}_{}_cg.pdb".format(file_number, file_type)
X_list, Y_list, Z_list, atomtype_list = read_pdb(
"./training_data/{}".format(file_name))
res.append([X_list, Y_list, Z_list, atomtype_list])
return res
def plot_distribution_range_xyz():
num = []
x = []
y = []
z = []
for i in range(1, 3001):
print(i)
X_list, Y_list, Z_list, atomtype_list = read_pdb(
'../data/training_data/%s_pro_cg.pdb' % str(i).zfill(4))
num.append((len(X_list)))
x.append(max(X_list) - min(X_list))
y.append(max(Y_list) - min(Y_list))
z.append(max(Z_list) - min(Z_list))
fig = plt.figure(dpi=128, figsize=(12, 6))
fig.autofmt_xdate()
fig.suptitle('Spatial Range VS Number of Atom within a Protein')
ax = plt.subplot(131)
ax.set_title('X Range VS Number of Atoms')
ax.set_ylabel('X')
ax.set_xlabel('The Numbers of Atoms within a Protein')
ax.scatter(num, x, label='Protein')
plt.legend()
plt.grid(True)
ax = plt.subplot(132)
ax.set_title('Y Range VS Number of Atoms')
ax.set_ylabel('Y')
ax.set_xlabel('The Numbers of Atoms within a Protein')
ax.scatter(num, y, label='Protein')
plt.legend()
plt.grid(True)
ax = plt.subplot(133)
ax.set_title('Z Range VS Number of Atoms')
ax.set_ylabel('Z')
ax.set_xlabel('The Numbers of Atoms within a Protein')
ax.scatter(num, z, label='Protein')
plt.legend()
plt.grid(True)
plt.tight_layout()
plt.subplots_adjust(top=0.88)
plt.show()
def extract_data(i, type):
x_list, y_list, z_list, atomtype_list = read_pdb('../data/training_data/%s_%s_cg.pdb' % (str(i).zfill(4), type))
data = np.transpose(np.array([x_list, y_list, z_list, atomtype_list]))
return np.array(data)