def _read_header(self):
"""Priveat method called by ``__init__`` to read all file headers into the class attributes and calculate
sequence dependant values.
:return: headers in class attributes.
"""
d = GlobalDescriptor('X') # template
# loop through all files in the directory
for i, file in enumerate(self.filenames):
with open(join(
self.directory,
file)) as f: # read first 4 lines as header, rest as data
head = [next(f) for _ in range(4)]
data = [next(f) for _ in range(4, (self.wmax - self.wmin) + 5)]
# read headers into class attributes
name = head[0].split('\r\n')[0]
self.names.append(name)
sequence = head[1].split('\r\n')[0].strip()
self.sequences.append(sequence)
umol = float(head[2].split('\r\n')[0])
self.conc_umol.append(umol)
self.solvent.append(head[3].split('\r\n')[0])
# read CD data
wlengths = [int(line.split(',')[0])
for line in data] # get rid of s***** line ends
ellipts = [
float(line.split(',')[1].split('\r\n')[0]) for line in data
]
self.circular_dichroism.append(
np.array(list(zip(wlengths, ellipts))))
# calculate MW and transform concentration to mg/ml
d.sequences = [sequence]
d.calculate_MW(amide=self.amide)
self.mw.append(d.descriptor[0][0])
self.conc_mgml.append(self.mw[i] * umol / 10**6)
self.meanres_mw.append(
self.mw[i] / (len(sequence) -
1)) # mean residue molecular weight (MW / n-1)
def predict():
if request.method == 'POST':
seq = request.form['seq']
with open("random.fasta", "w") as fp:
fp.write(seq)
pepdesc = PeptideDescriptor(
'/home/sanika/proj/random.fasta',
'eisenberg') # use Eisenberg consensus scale
globdesc = GlobalDescriptor('/home/sanika/proj/random.fasta')
# --------------- Peptide Descriptor (AA scales) Calculations ---------------
pepdesc.calculate_global() # calculate global Eisenberg hydrophobicity
pepdesc.calculate_moment(
append=True) # calculate Eisenberg hydrophobic moment
# load other AA scales
pepdesc.load_scale('gravy') # load GRAVY scale
pepdesc.calculate_global(
append=True) # calculate global GRAVY hydrophobicity
pepdesc.calculate_moment(
append=True) # calculate GRAVY hydrophobic moment
pepdesc.load_scale('z3') # load old Z scale
pepdesc.calculate_autocorr(
1,
append=True) # calculate global Z scale (=window1 autocorrelation)
# --------------- Global Descriptor Calculations ---------------
globdesc.length() # sequence length
globdesc.boman_index(append=True) # Boman index
globdesc.aromaticity(append=True) # global aromaticity
globdesc.aliphatic_index(append=True) # aliphatic index
globdesc.instability_index(append=True) # instability index
globdesc.calculate_charge(ph=7.4, amide=False,
append=True) # net charge
globdesc.calculate_MW(amide=False, append=True) # molecular weight
f1 = pepdesc.descriptor
f2 = globdesc.descriptor
result = np.concatenate((f2, f1), axis=1)
clf = joblib.load('ml_model.pkl')
pred = clf.predict(result)
proba = clf.predict_proba(result).tocoo()
mc = pred.tocoo()
out = mc.col
res = []
labels = ['antiviral', 'antibacterial', 'antifungal']
values = proba.data
plt.pie(values,
labels=labels,
autopct='%.0f%%',
shadow=True,
radius=0.5)
plt.savefig('/home/sanika/proj/pie_chart.jpg')
figfile = BytesIO()
plt.savefig(figfile, format='png')
figfile.seek(0)
figdata_png = base64.b64encode(figfile.getvalue()).decode('ascii')
plt.close()
for i in range(len(out)):
if out[i] == 0:
res.append("antiviral")
elif out[i] == 1:
res.append("antibacterial")
else:
res.append("antifungal")
return render_template('seq.html', seq=res, result=figdata_png)
return render_template('predictor.html')
def upload():
if request.method == 'POST':
# This will be executed on POST request.
upfile = request.files['file']
if upfile and allowed_file(upfile.filename):
filename = secure_filename(upfile.filename)
upfile.save(os.path.join(app.config['UPLOAD_FOLDER'], filename))
#return render_template('upload.html')
#flash("File uploaded", "success")
#with open("/home/sanika/proj/uploads/aa.fasta") as f:
#lines = f.readlines()
#lines = [l for l in lines if "ROW" in l]
#with open("/home/sanika/proj/uploads/out.fasta", "w") as f1:
#f1.writelines(lines)
#f = open(filename)
#prot_seq = ReadFasta(f)
with open(filename) as fasta_file: # Will close handle cleanly
identifiers = []
sequence = []
for seq_record in SeqIO.parse(fasta_file,
'fasta'): # (generator)
identifiers.append(seq_record.id)
sequence.append(seq_record.seq)
pepdesc = PeptideDescriptor(
filename, 'eisenberg') # use Eisenberg consensus scale
globdesc = GlobalDescriptor(filename)
# --------------- Peptide Descriptor (AA scales) Calculations ---------------
pepdesc.calculate_global(
) # calculate global Eisenberg hydrophobicity
pepdesc.calculate_moment(
append=True) # calculate Eisenberg hydrophobic moment
# load other AA scales
pepdesc.load_scale('gravy') # load GRAVY scale
pepdesc.calculate_global(
append=True) # calculate global GRAVY hydrophobicity
pepdesc.calculate_moment(
append=True) # calculate GRAVY hydrophobic moment
pepdesc.load_scale('z3') # load old Z scale
pepdesc.calculate_autocorr(
1, append=True
) # calculate global Z scale (=window1 autocorrelation)
# --------------- Global Descriptor Calculations ---------------
globdesc.length() # sequence length
globdesc.boman_index(append=True) # Boman index
globdesc.aromaticity(append=True) # global aromaticity
globdesc.aliphatic_index(append=True) # aliphatic index
globdesc.instability_index(append=True) # instability index
globdesc.calculate_charge(ph=7.4, amide=False,
append=True) # net charge
globdesc.calculate_MW(amide=False, append=True) # molecular weight
f1 = pepdesc.descriptor
f2 = globdesc.descriptor
result = np.concatenate((f2, f1), axis=1)
rs = []
for i in range(len(result)):
prt = np.reshape(result[i], (-1, 14))
clf = joblib.load('ml_model.pkl')
pred = clf.predict(prt)
out = pred.toarray()
#print(clf.predict_proba(result))
proba = clf.predict_proba(prt).tocoo()
mc = pred.tocoo()
out = mc.col
res = []
for i in range(len(out)):
if out[i] == 0:
res.append("antiviral")
elif out[i] == 1:
res.append("antibacterial")
else:
res.append("antifungal")
rs.append(res)
a = []
for i in range(len(rs)):
a.append('-'.join(rs[i]))
df = pd.DataFrame(data={
"id": identifiers,
"sequence": sequence,
"activity": a
},
columns=['id', 'sequence', 'activity'])
df.to_csv("result.csv", sep=',', index=False)
os.remove(os.path.join(app.config['UPLOAD_FOLDER'], filename))
#return render_template('seq.html', seq = rs)
return render_template('up.html', mimetype="text/csv")
#flash("File uploaded: Thanks!", "success")
else:
error = "PLEASE CHECK THE FORMAT OF FILE TO UPLOAD"
return render_template('upload.html', error=error)
# This will be executed on GET request.
return render_template('predictor.html')
globdesc = GlobalDescriptor('/path/to/sequences.fasta')
# --------------- Peptide Descriptor (AA scales) Calculations ---------------
pepdesc.calculate_global() # calculate global Eisenberg hydrophobicity
pepdesc.calculate_moment(append=True) # calculate Eisenberg hydrophobic moment
# load other AA scales
pepdesc.load_scale('gravy') # load GRAVY scale
pepdesc.calculate_global(append=True) # calculate global GRAVY hydrophobicity
pepdesc.calculate_moment(append=True) # calculate GRAVY hydrophobic moment
pepdesc.load_scale('z3') # load old Z scale
pepdesc.calculate_autocorr(
1, append=True) # calculate global Z scale (=window1 autocorrelation)
# save descriptor data to .csv file
col_names1 = 'ID,Sequence,H_Eisenberg,uH_Eisenberg,H_GRAVY,uH_GRAVY,Z3_1,Z3_2,Z3_3'
pepdesc.save_descriptor('/path/to/descriptors1.csv', header=col_names1)
# --------------- Global Descriptor Calculations ---------------
globdesc.length() # sequence length
globdesc.boman_index(append=True) # Boman index
globdesc.aromaticity(append=True) # global aromaticity
globdesc.aliphatic_index(append=True) # aliphatic index
globdesc.instability_index(append=True) # instability index
globdesc.calculate_charge(ph=7.4, amide=False, append=True) # net charge
globdesc.calculate_MW(amide=False, append=True) # molecular weight
# save descriptor data to .csv file
col_names2 = 'ID,Sequence,Length,BomanIndex,Aromaticity,AliphaticIndex,InstabilityIndex,Charge,MW'
globdesc.save_descriptor('/path/to/descriptors2.csv', header=col_names2)
desc.formula(amide=True)
for v in desc.descriptor:
formula_array.append(v[0])
except:
formula_array.append('')
database['formula'] = formula_array
print("Estimate molecular_weigth")
#get MW for each sequence
molecular_weigth_array = []
for i in range(len(database)):
try:
desc = GlobalDescriptor([database['Sequence'][i]])
desc.calculate_MW(amide=True)
molecular_weigth_array.append(desc.descriptor)
except:
molecular_weigth_array.append('')
database['molecular_weigth'] = molecular_weigth_array
print("Estimate charge")
#calculate charge for each sequence
charge_array = []
for i in range(len(database)):
try:
desc = GlobalDescriptor([database['Sequence'][i]])
desc.calculate_charge(ph=7, amide=True)
charge_array.append(desc.descriptor[0][0])
def exec(peptide, time_node):
file = open("../src/public/jobs/service1/service1.fasta", "w")
file.write(peptide)
file.close()
fasta = SeqIO.parse("../src/public/jobs/service1/service1.fasta", "fasta")
if(any(fasta) == False): #False when `fasta` is empty
return "error"
cantidad = 0
for record in SeqIO.parse("../src/public/jobs/service1/service1.fasta", "fasta"):
cantidad = cantidad+1
if (cantidad == 1):
properties = {}
for record in SeqIO.parse("../src/public/jobs/service1/service1.fasta", "fasta"):
properties[str(record.id)] = {}
#save properties
properties[str(record.id)]["length"] = len(record.seq)
#formula
try:
desc = GlobalDescriptor(str(record.seq))
desc.formula(amide=True)
properties[str(record.id)]["formula"] = desc.descriptor[0][0]
except:
properties[str(record.id)]["formula"] = "-"
#molecular weigth
try:
desc = GlobalDescriptor(str(record.seq))
desc.calculate_MW(amide=True)
properties[str(record.id)]["molecular_weigth"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["molecular_weigth"] = "-"
#boman_index
try:
desc = GlobalDescriptor(str(record.seq))
desc.boman_index()
properties[str(record.id)]["boman_index"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["boman_index"] = "-"
#charge
try:
desc = GlobalDescriptor(str(record.seq))
desc.calculate_charge(ph=7, amide=True)
properties[str(record.id)]["charge"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["charge"] = "-"
#charge density
try:
desc = GlobalDescriptor(str(record.seq))
desc.charge_density(ph=7, amide=True)
properties[str(record.id)]["charge_density"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["charge_density"] = "-"
#estimate isoelectric point
try:
desc = GlobalDescriptor(str(record.seq))
desc.isoelectric_point()
properties[str(record.id)]["isoelectric_point"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["isoelectric_point"] = "-"
#estimate inestability index
try:
desc = GlobalDescriptor(str(record.seq))
desc.instability_index()
properties[str(record.id)]["instability_index"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["instability_index"] = "-"
#estimate aromaticity
try:
desc = GlobalDescriptor(str(record.seq))
desc.aromaticity()
properties[str(record.id)]["aromaticity"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["aromaticity"] = "-"
#estimate aliphatic_index
try:
desc = GlobalDescriptor(str(record.seq))
desc.aliphatic_index()
properties[str(record.id)]["aliphatic_index"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["aliphatic_index"] = "-"
#estimate hydrophobic_ratio
try:
desc = GlobalDescriptor(str(record.seq))
desc.hydrophobic_ratio()
properties[str(record.id)]["hydrophobic_ratio"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["hydrophobic_ratio"] = "-"
#profile hydrophobicity
try:
desc = PeptideDescriptor(str(record.seq), scalename='Eisenberg')
desc.calculate_profile(prof_type='H')
properties[str(record.id)]["hydrophobicity_profile"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["hydrophobicity_profile"] = "-"
#profile hydrophobic
try:
desc = PeptideDescriptor(str(record.seq), scalename='Eisenberg')
desc.calculate_profile(prof_type='uH')
properties[str(record.id)]["hydrophobic_profile"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["hydrophobic_profile"] = "-"
#moment
try:
desc = PeptideDescriptor(str(record.seq), scalename='Eisenberg')
desc.calculate_moment()
properties[str(record.id)]["calculate_moment"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["calculate_moment"] = "-"
try:
os.mkdir("../src/public/jobs/service1/"+time_node)
except:
print("Error")
#generate plot profile
plot_profile(str(record.seq), scalename='eisenberg', filename= "../src/public/jobs/service1/"+time_node+"/profile.png")
#generate helical wheel
helical_wheel(str(record.seq), colorcoding='charge', lineweights=False, filename= "../src/public/jobs/service1/"+time_node+"/helical.png")
return(properties)
if (cantidad > 1):
properties = {}
for record in SeqIO.parse("../src/public/jobs/service1/service1.fasta", "fasta"):
properties[str(record.id)] = {}
properties[str(record.id)]["length"] = len(record.seq)
#formula
try:
desc = GlobalDescriptor(str(record.seq))
desc.formula(amide=True)
properties[str(record.id)]["formula"] = desc.descriptor[0][0]
except:
properties[str(record.id)]["formula"] = "-"
#molecular weigth
try:
desc = GlobalDescriptor(str(record.seq))
desc.calculate_MW(amide=True)
properties[str(record.id)]["molecular_weigth"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["molecular_weigth"] = "-"
#boman_index
try:
desc = GlobalDescriptor(str(record.seq))
desc.boman_index()
properties[str(record.id)]["boman_index"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["boman_index"] = "-"
#charge
try:
desc = GlobalDescriptor(str(record.seq))
desc.calculate_charge(ph=7, amide=True)
properties[str(record.id)]["charge"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["charge"] = "-"
#charge density
try:
desc = GlobalDescriptor(str(record.seq))
desc.charge_density(ph=7, amide=True)
properties[str(record.id)]["charge_density"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["charge_density"] = "-"
#estimate isoelectric point
try:
desc = GlobalDescriptor(str(record.seq))
desc.isoelectric_point()
properties[str(record.id)]["isoelectric_point"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["isoelectric_point"] = "-"
#estimate inestability index
try:
desc = GlobalDescriptor(str(record.seq))
desc.instability_index()
properties[str(record.id)]["instability_index"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["instability_index"] = "-"
#estimate aromaticity
try:
desc = GlobalDescriptor(str(record.seq))
desc.aromaticity()
properties[str(record.id)]["aromaticity"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["aromaticity"] = "-"
#estimate aliphatic_index
try:
desc = GlobalDescriptor(str(record.seq))
desc.aliphatic_index()
properties[str(record.id)]["aliphatic_index"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["aliphatic_index"] = "-"
#estimate hydrophobic_ratio
try:
desc = GlobalDescriptor(str(record.seq))
desc.hydrophobic_ratio()
properties[str(record.id)]["hydrophobic_ratio"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["hydrophobic_ratio"] = "-"
#profile hydrophobicity
try:
desc = PeptideDescriptor(str(record.seq), scalename='Eisenberg')
desc.calculate_profile(prof_type='H')
properties[str(record.id)]["hydrophobicity_profile"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["hydrophobicity_profile"] = "-"
#profile hydrophobic
try:
desc = PeptideDescriptor(str(record.seq), scalename='Eisenberg')
desc.calculate_profile(prof_type='uH')
properties[str(record.id)]["hydrophobic_profile"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["hydrophobic_profile"] = "-"
#moment
try:
desc = PeptideDescriptor(str(record.seq), scalename='Eisenberg')
desc.calculate_moment()
properties[str(record.id)]["calculate_moment"] = float("%.4f" % desc.descriptor[0][0])
except:
properties[str(record.id)]["calculate_moment"] = "-"
return(properties)
from modlamp.descriptors import GlobalDescriptor
sequences = []
MIC = []
units = []
actives = {}
# read the file with 3 columns containing MIC values
with open('Saureus.csv', 'r') as f:
for line in f:
sequences.append(line.split(',')[0])
MIC.append(line.split(',')[1])
units.append(line.split(',')[2])
D = GlobalDescriptor(sequences)
D.calculate_MW()
MW = D.descriptor.tolist()
for i, u in enumerate(units):
if u == 'ug/ml\r\n': # find MIC values in ug/mL
if '+' in MIC[i]:
mic = float(MIC[i].split('+')[0]) + float(MIC[i].split('+')[1]) # if with stdev, take upper bound
actives[sequences[i]] = round((mic / float(MW[i][0])) * 1000., 1) # convert ug/mL to uM
elif '-' in MIC[i]:
mic = float(MIC[i].split('-')[1]) # if with stdev, be conservative and take upper bound
actives[sequences[i]] = round((mic / float(MW[i][0])) * 1000., 1) # convert ug/mL to uM
else:
actives[sequences[i]] = round((float(MIC[i]) / float(MW[i][0])) * 1000., 1) # convert ug/mL to uM
s_inactive = [s for s, v in actives.items() if v > 100.0]
s_active = [s for s, v in actives.items() if v <= 100.0]