def dist_pI(output, plot_suffix, EC):
"""
Plot distribution of protein pI for all sequences in FASTA file.
"""
fig, ax = plt.subplots(figsize=(8, 5))
width = 0.5
X_bins = arange(0, 14.1, width)
hist, bin_edges = histogram(a=list(output.pI), bins=X_bins)
# output.GRAVY.plot.hist(bins=50,
# color='#607c8e')
ax.bar(bin_edges[:-1],
hist,
align='edge',
alpha=0.4,
width=width,
color=EC_descriptions()[str(EC)][1],
edgecolor='k',
label=EC_descriptions()[str(EC)][0])
# ax.plot(X_bins[:-1]+width/2, hist, c='k', lw='2')
dist_plot(fig,
ax,
name='pI',
xlim=(0, 14),
xtitle='pI',
plot_suffix=plot_suffix)
def dist_GRAVY(output, plot_suffix, EC):
"""
Plot distribution of protein GRAVY for all sequences in FASTA file.
"""
fig, ax = plt.subplots(figsize=(8, 5))
width = 0.05
X_bins = arange(-2, 2.2, width)
hist, bin_edges = histogram(a=list(output.GRAVY), bins=X_bins)
# output.GRAVY.plot.hist(bins=50,
# color='#607c8e')
ax.bar(bin_edges[:-1],
hist,
align='edge',
alpha=0.4,
width=width,
color=EC_descriptions()[str(EC)][1],
edgecolor='k',
label=EC_descriptions()[str(EC)][0])
# ax.plot(X_bins[:-1]+width/2, hist, c='k', lw='2')
# GRAVY specific visuals
# ax.text(-1.45, 40, 'hydrophilic', fontsize=16)
# get ylim
ylim = ax.get_ylim()
ax.text(0.55,
max(ylim) / 2 + 0.05 * (max(ylim) / 2),
'hydrophobic',
fontsize=16)
ax.arrow(0.5,
max(ylim) / 2,
0.7,
0,
head_width=0.05 * (max(ylim) / 2),
head_length=0.1,
fc='k',
ec='k')
# avg_GRAVY = -0.4
# ax.axvline(x=avg_GRAVY, c='grey', alpha=1.0, linestyle='--')
# catalase_GRAVY = -0.605
# ax.axvline(x=catalase_GRAVY, c='r', alpha=1.0)
# urease_GRAVY = -0.1524
# ax.axvline(x=urease_GRAVY, c='b', alpha=1.0)
dist_plot(fig,
ax,
name='GRAVY',
xlim=(-1.5, 1.5),
xtitle='GRAVY',
plot_suffix=plot_suffix)
def dist_Iindex(output, plot_suffix, EC):
"""
Plot distribution of protein I index for all sequences in
FASTA file.
"""
fig, ax = plt.subplots(figsize=(8, 5))
width = 5
X_bins = arange(0, 150, width)
hist, bin_edges = histogram(a=list(output.I_index), bins=X_bins)
# output.GRAVY.plot.hist(bins=50,
# color='#607c8e')
# ax.plot(X_bins[:-1]+width/2, hist, c='k', lw='2')
ax.bar(bin_edges[:-1],
hist,
align='edge',
alpha=0.4,
width=width,
color=EC_descriptions()[str(EC)][1],
edgecolor='k',
label=EC_descriptions()[str(EC)][0])
# instability specific visuals
# get ylim
ylim = ax.get_ylim()
ax.text(51,
max(ylim) / 2 + 0.05 * (max(ylim) / 2),
'unstable',
fontsize=16)
ax.arrow(50,
max(ylim) / 2,
30,
0,
head_width=0.05 * (max(ylim) / 2),
head_length=4,
fc='k',
ec='k')
II_cutoff = 40
ax.axvline(x=II_cutoff, c='k', alpha=1.0, linestyle='--', lw=2)
# catalase_II = 27.010
# ax.axvline(x=catalase_II, c='r', alpha=1.0)
# urease_II = 31.75
# ax.axvline(x=urease_II, c='b', alpha=1.0)
dist_plot(fig,
ax,
name='Iindex',
xlim=(0, 100),
xtitle='instability index',
plot_suffix=plot_suffix)
def main_analysis(plot_suffix, fasta_file, output_file, EC):
"""Analyse all sequences in FASTA file from BRENDA.
"""
print('---------------------------------------------------------')
print('Analyse properties of all sequences in FASTA:', fasta_file)
print('---------------------------------------------------------')
# percent_w_sequence(output_dir=search_output_dir)
temp_time = time.time()
if input('run calculations? (t/f)') == 't':
get_fasta_sequence_properties(output_file=output_file,
fasta_file=fasta_file)
# do plotting + analysis -- plots
if EC in list(EC_descriptions().keys()):
print('------------------------------------------------------')
print('doing analysis...')
# load existing data from this FASTA file
fasta_plotting(output_file=output_file, plot_suffix=plot_suffix, EC=EC)
print('--- time taken =', '{0:.2f}'.format(time.time() - temp_time),
's')
else:
print('------------------------------------------------------')
print('doing specific sequence analysis...')
output = read_seq_output(output_file)
dist_TMindex_specific(output, plot_suffix, EC)
def screen_pIs(database_names, redo_pI, redo_pI_plots, pI_csv, pI_output_dir,
cutoff_pi, descriptors):
"""
Screen the pI of all sequences with chosen EC numbers.
"""
if descriptors is None:
descriptors = {}
for EC_file in database_names:
EC = EC_file.replace(pI_output_dir, '')
EC = EC.replace('__BRENDA_sequences.fasta', '').replace('_', '.')
top_EC = EC.split('.')[0]
# read the file but to avoid memory issues # we will calculate
# the pI on the fly using the bio python module
print('doing:', EC_file)
file_mod = EC_file.replace(".fasta", "_mod.fasta")
if redo_pI is True:
calculate_pI_from_file(file_mod, pI_output_dir, cutoff_pi, pI_csv)
if redo_pI_plots is True:
print('plot distribution of pIs')
pi_data = pd.read_csv(pI_output_dir + pI_csv, index_col=False)
EC_pi_data = pi_data[pi_data['fasta_file'] == file_mod]
plot_EC_pI_dist(EC_pi_data,
filename=file_mod.replace('.fasta', '.pdf'),
title=EC_descriptions()[top_EC][0],
cutoff_pi=cutoff_pi)
print('done')
if redo_pI_plots is True:
print('plot full distribution of pIs')
pi_data = pd.read_csv(pI_output_dir + pI_csv, index_col=False)
plot_pI_dist(pi_data,
filename='full_pI_dist.pdf',
output_dir=pI_output_dir,
cutoff_pi=cutoff_pi)
def dist_Aindex(output, plot_suffix, EC):
"""
Plot distribution of protein Aindex for all sequences in FASTA file.
"""
fig, ax = plt.subplots(figsize=(8, 5))
width = 5
X_bins = arange(0, 150, width)
hist, bin_edges = histogram(a=list(output.A_index), bins=X_bins)
# output.GRAVY.plot.hist(bins=50,
# color='#607c8e')
# ax.plot(X_bins[:-1]+width/2, hist, c='k', lw='2')
ax.bar(bin_edges[:-1],
hist,
align='edge',
alpha=0.4,
width=width,
color=EC_descriptions()[str(EC)][1],
edgecolor='k',
label=EC_descriptions()[str(EC)][0])
# AI specific visuals
ylim = ax.get_ylim()
ax.text(10,
max(ylim) / 2 + 0.05 * (max(ylim) / 2),
'more stable',
fontsize=16)
ax.arrow(10,
max(ylim) / 2,
40,
0,
head_width=0.05 * (max(ylim) / 2),
head_length=5,
fc='k',
ec='k')
# catalase_AI = 68
# ax.axvline(x=catalase_AI, c='r', alpha=1.0)
# urease_AI = 90.476
# ax.axvline(x=urease_AI, c='b', alpha=1.0)
dist_plot(fig,
ax,
name='Aindex',
xlim=(0, 150),
xtitle='aliphatic index',
plot_suffix=plot_suffix)
def dist_TMindex(output, plot_suffix, EC):
"""
Plot distribution of protein TM index for all sequences in
FASTA file.
"""
fig, ax = plt.subplots(figsize=(8, 5))
width = 0.2
X_bins = arange(-5, 5.1, width)
hist, bin_edges = histogram(a=list(output.TM_index), bins=X_bins)
# output.GRAVY.plot.hist(bins=50,
# color='#607c8e')
# ax.plot(X_bins[:-1]+width/2, hist, c='k', lw='2')
ax.bar(bin_edges[:-1],
hist,
align='edge',
alpha=0.4,
width=width,
color=EC_descriptions()[str(EC)][1],
edgecolor='k',
label=EC_descriptions()[str(EC)][0])
# melting temperature index specific visuals
TM_cutoff = (0, 1)
ax.axvspan(xmin=TM_cutoff[0],
xmax=TM_cutoff[1],
facecolor='grey',
alpha=0.2)
# catalase_TMI = 1.22
# ax.axvline(x=catalase_TMI, c='r', alpha=1.0)
# urease_TMI = 0.62
# ax.axvline(x=urease_TMI, c='b', alpha=1.0)
dist_plot(fig,
ax,
name='TMindex',
xlim=(-5, 5),
xtitle='thermostability index',
plot_suffix=plot_suffix)
def all_EC_violin_plot():
"""Do violin plots of all properties for all EC output files.
"""
properties = ['I_index', 'A_index', 'TM_index', 'pI', 'GRAVY']
prop_label = [
'instability index', 'aliphatic index', 'TM index', 'pI', 'GRAVY'
]
prop_lim = [(0, 100), (0, 150), (-5, 5), (0, 14), (-1.5, 1.5)]
ECs = ['1', '2', '3', '4', '5', '6']
output_files = [i + '__BRENDA_sequences_output.csv' for i in ECs]
for i, prop in enumerate(properties):
print('doing', prop, '....')
fig, ax = plt.subplots(figsize=(8, 5))
for out_file in output_files:
print(out_file)
EC = out_file[0]
print(EC)
output = read_seq_output(out_file)
parts = ax.violinplot(
output[prop],
[int(EC)],
showmeans=False,
showmedians=False,
showextrema=False,
)
for pc in parts['bodies']:
pc.set_facecolor(EC_descriptions()[EC][1])
pc.set_edgecolor('black')
pc.set_alpha(0.6)
if prop == 'TM_index':
# melting temperature index specific visuals
TM_cutoff = (0, 1)
ax.axhspan(ymin=TM_cutoff[0],
ymax=TM_cutoff[1],
facecolor='grey',
alpha=0.2)
if prop == 'I_index':
II_cutoff = 40
ax.axhline(y=II_cutoff, c='k', alpha=1.0, linestyle='--', lw=2)
if prop == 'A_index':
ax.text(0.21,
60,
'more stable',
fontsize=16,
ha='left',
va='bottom',
rotation=90)
ax.arrow(0.5,
40,
0,
80,
head_width=0.2,
head_length=10,
fc='k',
ec='k')
ax.tick_params(axis='both', which='major', labelsize=16)
ax.set_xlabel('EC number', fontsize=16)
ax.set_ylabel(prop_label[i], fontsize=16)
ax.set_xlim(0, 7)
ax.set_ylim(prop_lim[i])
ax.set_xticks([1, 2, 3, 4, 5, 6])
ax.set_xticklabels(['1', '2', '3', '4', '5', '6'])
fig.tight_layout()
fig.savefig("violin_" + prop + ".pdf", dpi=720, bbox_inches='tight')