def test_create_networkx_graph_duplicates(self):
networks_table = pd.DataFrame({'GUID' : ['guid'], 'Name': ['net']})
edge_table = pd.DataFrame({'Source': ['1', '1'], 'Target': ['2', '2']})
node_table = pd.DataFrame({'Node': ['1', '1', '2']})
import warnings
with warnings.catch_warnings(record=True) as w:
graphs = nx.from_perseus(networks_table, {'guid': {'name' : 'net', 'guid': 'guid', 'edge_table': edge_table, 'node_table': node_table}})
self.assertEqual(2, graphs[0].number_of_nodes())
self.assertEqual(1, graphs[0].number_of_edges())
warning_messages = [str(x.message) for x in w if 'deprecated' not in str(x.message)]
self.assertTrue('Duplicate edges' in warning_messages[0], warning_messages[0])
self.assertTrue('Duplicate nodes' in warning_messages[1], warning_messages[1])
def test_write_bool_column_as_categorical(self):
df = pd.DataFrame(columns=pd.Index(['Significant'], name='Column Name'))
df['Significant'] = [True, False, True, True]
self.assertEqual(df.dtypes[0], np.dtype('bool'))
df_str = to_string(df)
self.assertEqual('Significant\n#!{Type}C\n+\n""\n+\n+\n', df_str, df_str)
df_str = to_string(df, convert_bool_to_category=False)
self.assertEqual('Significant\n#!{Type}C\nTrue\nFalse\nTrue\nTrue\n', df_str, df_str)
def test_inferring_and_setting_main_columns(self):
df = pd.DataFrame({
'a': [2, 3],
'b': [1, 2],
'c': ['a', 'b'],
'd': [3, 4]
})
self.assertEqual('#!{Type}E\tE\tT\tN', type_row(df))
self.assertEqual('#!{Type}N\tE\tT\tE',
type_row(df, main_columns={'b', 'd'}))
def countAA(maxcluster):
dpAAs = pd.DataFrame(['nterm','cterm','G','A','V','L','I','P','F','Y','W','S','T',\
'C','M','N','Q','K','R','H','D','E'])
dpAAs['Count'] = 0
for x in range(len(maxcluster) - 1):
x = x + 1
i = dpAAs['Count'][dpAAs[0].isin(list(maxcluster['DP AA'][x]\
.split(';')))].index
dpAAs['Count'].loc[i] = dpAAs['Count'].loc[i] + 1
return dpAAs
def pltcountperRaw():
x = int(input('How many files do you want to plot together? '))
a = openfile()
withoutunmodified = a[a['DP Modification'] == 'Unmodified']
unmod = pd.DataFrame(withoutunmodified['Raw file'].value_counts())
values = pd.DataFrame(a['Raw file'].value_counts())
# =============================================================================
# #just for MSFragger files
# b = openfile()
# b['Raw file'] = b['Spectrum'].apply(lambda df: df.split('.')[0])
# b = b[b['Observed Modifications']!='Unknown']
# values['MSFragger identifications'] = b['Raw file'].value_counts().values
# =============================================================================
for y in range(x - 1):
a = openfile()
withoutunmodified = a[a['DP Modification'] == 'Unmodified']
values['File ' + str(y)] = a['Raw file'].value_counts().values
unmod['File ' +
str(y)] = withoutunmodified['Raw file'].value_counts().values
values = values.sort_index()
unmod = unmod.sort_index()
values = values.reset_index(drop=True)
unmod = unmod.reset_index(drop=True)
if x == 3:
values = values.rename(index = str, columns = {'Raw file':'No matching',\
'File 0':'+/-1 fraction matching',\
'File 1':'Complete matching'})
unmod = unmod.rename(index = str, columns = {'Raw file':'No matching',\
'File 0':'Matching restricted to one raw file',\
'File 1':'Matching between all raw files'})
ax = values.plot.bar(rot=0, fontsize=20)
#unmod.plot.bar(ax=ax,color=['cornflowerblue', 'orange', 'limegreen'],fontsize = 15,legend = False,rot=0)
ax.legend(fontsize=20)
ax.set_xticklabels(list(range(1, len(values) + 1)))
ax.set_xlabel('Raw file', fontsize=20)
ax.set_ylabel('Count', fontsize=20)
return unmod
def findallin (cache,position,thestr,char1,char2):
# initialize cache,postiton as []
if(thestr.find(char1) > -1 and thestr.find(char2) > -1):
substr = thestr[thestr.find(char1)+1:thestr.find(char2)]
subsuperstr = thestr[thestr.find(char2)+1:len(thestr)]
cache.append(substr)
if not position:
position.append(thestr.find(char1)-1)
else:
idex = position[len(position)-1] + thestr.find(char1)
position.append(idex)
findallin(cache,position,subsuperstr,char1,char2)
df = pd.DataFrame(cache)
df.insert(1,1,position)
return df
def countAA(maxcluster):
# to do: include nterm, cterm here:
dpAAs = pd.DataFrame(['nterm','cterm','G','A','V','L','I','P','F','Y','W','S','T',\
'C','M','N','Q','K','R','H','D','E'])
dpAAs['Count'] = 0
dpAAs['nterm'] = 0
dpAAs['middle'] = 0
dpAAs['cterm'] = 0
for x in range(len(maxcluster) - 1):
x = x + 1
i = dpAAs['Count'][dpAAs[0].isin(list(maxcluster['DP AA'][x]\
.split(';')))].index
dpAAs['Count'].loc[i] = dpAAs['Count'].loc[i] + 1
if i[0] == 0:
dpAAs['nterm'].loc[i[1]] = dpAAs['nterm'].loc[i[1]] + 1
elif i[0] == 1:
dpAAs['cterm'].loc[i[1]] = dpAAs['cterm'].loc[i[1]] + 1
else:
dpAAs['middle'].loc[i] = dpAAs['middle'].loc[i] + 1
return dpAAs
parameters = parse_parameters(paramfile) # parse the parameters file
df = pd.read_perseus(infile) # read the input matrix into a pandas.DataFrame
n_neighbor = intParam(parameters, "Number of neighbors")
n_component = intParam(parameters, "Number of components")
seed = intParam(parameters, "Random state")
m_dist = doubleParam(parameters, "Minimum distance")
metric = singleChoiceParam(parameters, "Metric")
annotations = read_annotations(infile)
if len(annotations) < 2:
sys.exit("The data needs to be grouped")
newDF1 = main_df(infile, df)
newDF1 = newDF1.T
embedding = umap.UMAP(n_neighbors=n_neighbor,
n_components=n_component,
metric=metric,
random_state=seed,
min_dist=m_dist).fit_transform(newDF1)
new_annotations = {}
check_c = {}
c_num = 1
for k, v in annotations.items():
if "C:" in k:
col_n = k.replace("C:", "")
new_annotations[col_n] = v
annotation_df = pd.DataFrame.from_dict(new_annotations)
col_names = []
for i in range(0, n_component):
col_names.append("Component " + str(i + 1))
newDF2 = pd.DataFrame(data=embedding, columns=col_names)
newDF2 = pd.concat([newDF2.reset_index(drop=True), annotation_df], axis=1)
newDF2.to_perseus(outfile) # write pandas.DataFrame in Perseus txt format
def test_writing_empty_table_should_have_all_columns(self):
df = pd.DataFrame(columns=pd.Index(['Node'], name='Column Name'))
self.assertEqual(1, len(df.columns))
self.assertEqual('Column Name', df.columns.name)
df_str = to_string(df)
self.assertEqual('Node\n#!{Type}T\n', df_str, df_str)
def testMQ():
root = Tk()
root.withdraw()
a = filedialog.askopenfilename(initialdir = "D:\Thomas\PhytonSCripts\MQOutput" \
,title = "Choose a MQ nomatch.deppep file to plot",\
filetypes = (("deppep files","*.deppep"),("all files","*.*")))
a = pd.read_table(a, low_memory=False)
a = a.drop(0).reset_index(drop=True)
b = filedialog.askopenfilename(initialdir = "D:\Thomas\PhytonSCripts\MQOutput" \
,title = "Choose a MQ matching.deppep file to plot",\
filetypes = (("deppep files","*.deppep"),("all files","*.*")))
b = pd.read_table(b, low_memory=False)
b = b.drop(0).reset_index(drop=True)
arg = a[[
'Raw file', 'DP Base Raw File', 'DP Proteins', 'DP Base Sequence'
]]
argb = b[[
'Raw file', 'DP Base Raw File', 'DP Proteins', 'DP Base Sequence'
]]
uniq = arg['DP Base Sequence'].unique()
uniqb = argb['DP Base Sequence'].unique()
out = uniq[np.in1d(uniq, uniqb)]
df = pd.DataFrame({'Raw Files': arg['Raw file'].unique()})
df['counta'] = 0
df['countb'] = 0
count = 0
countb = 0
for x in range(len(out)):
coin = arg[arg['DP Base Sequence'] == out[x]]
coinb = argb[argb['DP Base Sequence'] == out[x]]
coincount = np.in1d(coin['Raw file'].unique(),
coinb['Raw file'].unique())
coinbcount = np.in1d(coinb['Raw file'].unique(),
coin['Raw file'].unique())
count = count + coincount[coincount == False].size
countb = countb + coinbcount[coinbcount == False].size
for y in (coin['Raw file'].unique()[coincount == False]):
df['counta'][df['Raw Files'] ==
y] = df['counta'][df['Raw Files'] == y] + 1
for z in (coinb['Raw file'].unique()[coinbcount == False]):
df['countb'][df['Raw Files'] ==
z] = df['countb'][df['Raw Files'] == z] + 1
df.plot.bar()
print('Amount of peptide sequences found in the Raw Files of file 1 but not'\
' found in the Raw Files of file 2 is ',count, '. For file 2: ', countb)
print('!!!Both peptide sequences need to be present in both files!!!')
print('Amount of dp. Peptides in file 1: ', len(a), ' in file 2: ', len(b))
values = pd.DataFrame(a['Raw file'].value_counts())
values['Raw'] = b['Raw file'].value_counts().values
values.columns = ['No match', 'Matching']
values = values.sort_index()
values.plot.bar()
a['Intensity'] = a['Intensity'].astype(float)
b['Intensity'] = b['Intensity'].astype(float)
raw_tablea = pd.pivot_table(a, values = 'Intensity',\
index = ['DP Cluster Index','DP AA','DP Base Sequence','DP Probabilities'], columns = 'Raw file')
raw_tableb = pd.pivot_table(b, values = 'Intensity',\
index = ['DP Cluster Index','DP AA','DP Base Sequence','DP Probabilities'], columns = 'Raw file')
raw_tablea = raw_tablea.reset_index()
raw_tableb = raw_tableb.reset_index()
return raw_tablea, raw_tableb
def plotsMQ(mq_dep, range_int, zerovar, minbin, binsize, binrange):
# Pie Plot of MQ found Modifications
print('MaxQuant Plot Parameters: min. modification: '+ str(range_int) + \
'; removed '+str(zerovar)+ 'bins around zero'+'; min. binsize : '\
+str(minbin)+'; binsize: '+str(binsize))
plt.figure()
dfMQ = mq_dep.groupby('DP Modification')\
.size().reset_index(name='No: of Modifications')
#make range of intrest relative to datasize?
range_intrest = range_int
df2 = dfMQ[dfMQ['No: of Modifications'] > range_intrest]
df3 = dfMQ[dfMQ['No: of Modifications'] < range_intrest]
others = 0
df2 = df2.reset_index(drop=True)
df3 = df3.reset_index(drop=True)
for x in range(len(df3)):
others = others + df3['No: of Modifications'][x]
df2.loc[len(df2)] = ['Others', others]
df2.set_index('DP Modification', inplace=True)
#df2.drop('Unknown', inplace=True) #dont kick them out
df2 = df2['No: of Modifications'].sort_values()
df2.plot.pie(y='No: of Modifications', legend=False).set_ylabel('')
# Bar PLot for MQ Mass Bins
plt.figure()
stepsize = binsize
#adjust binrange here
bins = np.arange(-binrange, binrange, stepsize)
var = round(len(bins) / 2)
sorted = round(mq_dep.sort_values('DP Cluster Mass'))
sorted['DP Cluster Mass'] = round(sorted['DP Cluster Mass'].astype(float))
sorted = sorted['DP Cluster Mass'].values
hist, bin_edges = np.histogram(sorted, bins=bins)
hist[var] = 0
# remove unknown modifications around 0
# adjusted Delta mass
# --> prob. no modification
#hist[hist>0] = np.log2(hist[hist>0])
plt.xlabel('Mass Bins', fontsize=15)
plt.ylabel('log-scale quantity', fontsize=13)
plt.bar(bin_edges[:-1], hist, width=1, log=True)
# plt.ylim([0,np.])
plt.show()
# Pie Plot for Bins
hist = np.append(hist, 0)
plt.figure()
histMQ = {'Histodata': hist, 'Bins': bins}
dMQ = pd.DataFrame(histMQ)
d2 = dMQ[dMQ['Histodata'] > minbin]
round_bin = np.around(d2['Bins'], decimals=2)
round_bin = round_bin.reset_index(drop=True)
for x in range(len(round_bin)):
round_bin[x] = str(round_bin[x])\
+ ' to ' + str(round((round_bin[x] + stepsize),2)) + ' bin'
d2.set_index(round_bin, inplace=True)
d2 = d2['Histodata'].sort_values()
d2.plot.pie(y='Histodata', legend=False).set_ylabel('')
return dMQ
def plotsMS(ms_psmopen, range_int, zerovar, minbin, binsize, binrange, nomi):
# Pie plot
print('MSFragger Plot Parameters: min. modification: '+ str(range_int) + \
'; removed '+str(zerovar)+ 'bins around zero'+'; min. binsize: '\
+str(minbin)+'; binsize: '+str(binsize))
plt.figure()
dfMS = ms_psmopen.groupby('Observed Modifications')\
.size().reset_index(name='No: of Modifications')
range_intrest = range_int
df2 = dfMS[dfMS['No: of Modifications'] > range_intrest]
df3 = dfMS[dfMS['No: of Modifications'] < range_intrest]
others = 0
df2 = df2.reset_index(drop=True)
df3 = df3.reset_index(drop=True)
for x in range(len(df3)):
others = others + df3['No: of Modifications'][x]
df2.loc[len(df2)] = ['Others', others]
df2.set_index('Observed Modifications', inplace=True)
df2.drop('Unknown', inplace=True)
df2 = df2['No: of Modifications'].sort_values()
df2.plot.pie(y='No: of Modifications', legend=False).set_ylabel('')
# Create new Bar Plot for Mass Bins ??not accurate == WHY??
plt.figure()
stepsize = binsize
# set x bin range here
bins = np.arange(-binrange, binrange, stepsize)
var = round(len(bins) / 2)
if (nomi == True):
sorted = round(ms_psmopen.sort_values('Adjusted Delta Mass'))
else:
sorted = ms_psmopen.sort_values('Adjusted Delta Mass')
hist, bin_edges = np.histogram(sorted['Adjusted Delta Mass'].values,
bins=bins)
hist[(var)] = 0 # remove unknown modifications around 0
# adjusted Delta mass
# --> prob. no modification
#hist[hist>0] = np.log2(hist[hist>0])
plt.xlabel('Mass Bins', fontsize=15)
plt.ylabel('log-scale quantity', fontsize=13)
plt.bar(bin_edges[:-1], hist, log=True, width=1)
# plt.ylim([0,np.log(10000000000000)])
plt.show()
# Pie PLot of Mass Bins
# do only once per program execution
hist = np.append(hist, 0)
plt.figure()
histMS = {'Histodata': hist, 'Bins': bins}
dMS = pd.DataFrame(histMS)
d2 = dMS[dMS['Histodata'] > minbin]
round_bin = np.around(d2['Bins'], decimals=2)
round_bin = round_bin.reset_index(drop=True)
for x in range(len(round_bin)):
round_bin[x] = str(round_bin[x])\
+ ' to ' + str(round((round_bin[x] + stepsize),2)) + ' bin'
d2.set_index(round_bin, inplace=True)
d2 = d2['Histodata'].sort_values()
d2.plot.pie(y='Histodata', legend=False).set_ylabel('')
return dMS