def getIndexChar(char):
import aaindex
aaindex.init(path='.')
dic = aaindex._aaindex.values()
listIndices = []
for item in dic:
# print(item.key)
listIndices.append(item.key)
listIndices = [
'KARS160114', 'KARS160106', 'MIYS990101', 'CASG920101', 'FAUJ880111',
'FAUJ880112'
]
listComputedValue = []
# forbiddenList=['KARS160114','TANS760101','MONM990201','NADH010101']
for indice in listIndices:
# if indice not in forbiddenList:
try:
x = aaindex.get(indice)
listComputedValue.append(x.get(char))
print(indice)
except Exception as e:
# print(indice)
# print("")
# l=2
print("ECHEC AVEC INDICES : " + indice)
return listComputedValue
def get_aaindex_features(AA):
"""
given an amino acid return its physicochemical features
"""
aaindex.init(path='../data/aaindex')
Hydrophobicity = aaindex.get('ARGP820101')
Hydrophilicity = aaindex.get('HOPT810101')
Polarity = aaindex.get('GRAR740102')
Polarizability = aaindex.get('CHAM820101')
Propensity = aaindex.get('WERD780101')
Average_accessible_surface_area = aaindex.get('JANJ780101')
Radius_of_gyration_of_side_chain = aaindex.get('LEVM760105')
Side_chain_volume = aaindex.get('KRIW790103')
Charge = aaindex.get('KLEP840101')
Number_of_hydrogen_bond_donors = aaindex.get('FAUJ880109')
Molecular_weight = aaindex.get('FASG760101')
Electron_ion_interaction_potential = aaindex.get('VELV850101')
#code to get features
aaindex_feature = [Hydrophobicity.get(AA),\
Hydrophilicity.get(AA),\
Polarity.get(AA),\
Polarizability.get(AA),\
Propensity.get(AA),\
Average_accessible_surface_area.get(AA),\
Radius_of_gyration_of_side_chain.get(AA),\
Side_chain_volume.get(AA),\
Charge.get(AA),\
Number_of_hydrogen_bond_donors.get(AA),\
Molecular_weight.get(AA),\
Electron_ion_interaction_potential.get(AA)]
return(aaindex_feature)
def _load_score_tab(self, **kwargs):
aaindex_dir = kwargs['aaindex_dir'] if kwargs.has_key('aaindex_dir') else 'datasets/aaindex/'
aaindex.init(aaindex_dir, '1')
scoretab = pd.DataFrame(index=list(Const.AMINO_ACIDS), )
for key, rec in aaindex._aaindex.iteritems():
# rec = aaindex.get(key)
vals = [rec.get(aa) for aa in Const.AMINO_ACIDS]
scoretab[key] = vals
# print('%s==>%s' % (key, rec.desc))
self._scoretab = scoretab
def getIndexSeq(seq):
import aaindex
aaindex.init(path='.')
dic = aaindex._aaindex.values()
listIndices = getIndicesUsed()
listComputedValue = [0] * len(listIndices)
forbiddenList = ['KARS160114', 'TANS760101', 'MONM990201', 'NADH010101']
for char in seq:
for i in range(len(listIndices)):
indice = listIndices[i]
try:
x = aaindex.get(indice)
listComputedValue[i] += (x.get(char))
except Exception as e:
print("ECHEC AVEC INDICES : " + indice)
return listComputedValue
def encode_aaindex_features(sequences):
aaindex.init(path='.', index='1')
aai_recs = [aaindex.get(d) for d in HQI8_descriptors]
return np.array([[get_aaindex_feature(aa, r, seq) for aa in seq
for r in aai_recs] for seq in sequences], dtype=np.float_)
#_author: Julian Zubek
import numpy as np
import aaindex
aaindex.init(path='.', index='1')
HQI8_descriptors = ["BLAM930101", "BIOV880101", "MAXF760101", "TSAJ990101",
"NAKH920108", "CEDJ970104", "LIFS790101", "MIYS990104"]
def get_aaindex_feature(amino_acid, aai_record, seq):
# B and Z are ambiguous amino acids.
if amino_acid == "B":
val = (aai_record.get("D") + aai_record.get("N")) / 2
elif amino_acid == "Z":
val = (aai_record.get("E") + aai_record.get("Q")) / 2
elif amino_acid == "O":
val = aai_record.get("K")
elif amino_acid == "U":
val = aai_record.get("C")
elif amino_acid in "X*-":
val = 0.0
else:
val = aai_record.get(amino_acid)
# Checking for "None" type in case of an unspecified amino acid character.
if not isinstance(val, float):
print("""Unrecognised amino acid symbol {0} found in sequence {1}
for descriptor {2}""".format(amino_acid, seq, aai_record))
exit(-1)
return val
# -*- coding: utf-8 -*-
"""
Created on Wed Dec 20 18:07:22 2017
@author: Patrick
"""
from __future__ import division
import aaindex
from aa_indeces import aai_to_get
from math import cos, sin, radians, degrees
aaindex.init(path='.')
def aaf(sequence, identifier, mode):
x = aaindex.get(identifier)
total = 0
total_list = []
for aa in sequence:
total = total + x.get(aa)
total_list.append(x.get(aa))
if mode == "mean":
return total / len(sequence)
elif mode == "total":
return total
elif mode == "max":
return max(total_list)
def aaf_angular(sequence, identifier, mode):
def get_vector(structure,\
path_file_pssm,\
path_aaindex,\
path_file_rsa,\
path_file_asa,\
list_dssp_features):
"""
get the features vector for a structure, returna list of all the features
for each residue
"""
aaindex.init(path_aaindex)
Hydrophobicity = aaindex.get('ARGP820101')
Hydrophilicity = aaindex.get('HOPT810101')
Polarity = aaindex.get('GRAR740102')
Polarizability = aaindex.get('CHAM820101')
Propensity = aaindex.get('WERD780101')
Average_accessible_surface_area = aaindex.get('JANJ780101')
Radius_of_gyration_of_side_chain = aaindex.get('LEVM760105')
Side_chain_volume = aaindex.get('KRIW790103')
Charge = aaindex.get('KLEP840101')
Number_of_hydrogen_bond_donors = aaindex.get('FAUJ880109')
Molecular_weight = aaindex.get('FASG760101')
Electron_ion_interaction_potential = aaindex.get('VELV850101')
list_vector = []
#asa
list_asa_value = get_asa_value(path_file_asa)
#rsa
list_rsa_value = get_rsa_relative(path_file_rsa)
#pssm
list_ipp_value, list_rwgrmtp_value = get_pssm_value(path_file_pssm)
#QIPI
QIPI = {
'H':1.147, 'R':1.346, 'K':0.784, 'A':0.841, 'V':0.994, 'I':1.084, 'L':1.144, \
'M':1.451, 'P':1.109, 'F':1.334, 'W':1.284, 'Y':1.368, 'G':0.823, 'C':1.172, \
'S':0.873, 'T':0.966, 'N':0.958, 'Q':0.909, 'D':0.830, 'E':0.805}
ppb = PPBuilder()
for pp in ppb.build_peptides(structure):
sequence = list(pp.get_sequence())
for i in range(len(sequence)):
list_vector.append([\
#asa
list_asa_value[i],\
#rsa
list_rsa_value[i],\
#pssm
list_ipp_value[i],\
list_rwgrmtp_value[i],\
#QIPI
QIPI[sequence[i]],\
#pseudo hydrophobicity
get_pseudo_hydrophobicity(sequence[i],\
Hydrophobicity,\
Charge),\
#aaindex
Hydrophobicity.get(sequence[i]),\
Hydrophilicity.get(sequence[i]),\
Polarity.get(sequence[i]),\
Polarizability.get(sequence[i]),\
Propensity.get(sequence[i]),\
Average_accessible_surface_area.get(sequence[i]),\
Radius_of_gyration_of_side_chain.get(sequence[i]),\
Side_chain_volume.get(sequence[i]),\
Charge.get(sequence[i]),\
Number_of_hydrogen_bond_donors.get(sequence[i]),\
Molecular_weight.get(sequence[i]),\
Electron_ion_interaction_potential.get(sequence[i])\
] + list_dssp_features[i])
#
return(list_vector)
from db_mapping.up_parsing import UPParser
hqi_db = "hqi.sqlite"
# liu_indices: hydrophobicity, hydrophicility, volumes of side chains of amino acids, polarity, polarizability, solvent-accessible surface area, net charge index
liu_indices = [
"ARGP820101", "HOPT810101", "KRIW790101", "GRAR740102", "CHAM820101",
"ROSG850101", "KLEP840101"
]
hqi8_indices = [
"BLAM930101", "BIOV880101", "MAXF760101", "TSAJ990101", "NAKH920108",
"CEDJ970104", "LIFS790101", "MIYS990104"
]
aaindex.init(path='.', index='1')
aai_recs = [aaindex.get(d) for d in liu_indices]
hqi8_recs = [aaindex.get(d) for d in hqi8_indices]
def get_aaindex_feature(amino_acid, aai_record):
# B and Z are ambiguous amino acids.
if amino_acid == "B":
val = (aai_record.get("D") + aai_record.get("N")) / 2
elif amino_acid == "Z":
val = (aai_record.get("E") + aai_record.get("Q")) / 2
elif amino_acid == "O":
val = aai_record.get("K")
elif amino_acid == "U":
val = aai_record.get("C")
elif amino_acid in "X*-":
import pandas as pd
import numpy as np
import pickle
import aaindex
from Bio import SeqIO
aaindex.init("./aaindex/")
def comp_sites(a, b, idx):
"""
Calculate dissimilarity between amino acids from AAIndex
"""
x = aaindex.get(idx)
m = x.get(a, a)
n = x.get(b, b)
k = x.get(a, b)
NoneType = type(None)
if isinstance(m, NoneType) or isinstance(n, NoneType) or isinstance(
k, NoneType):
# print "Unattended gaps in this AAIndex matrice"
dist = None
else:
dist = m + n - 2 * k
return dist
def pair_idx_encode(s1, s2, idx):
"""
Calculate dissimilarity between two sequences
from encoding_pairs import batch_idxList_encode
import pickle
import aaindex
aaindex.init('./aaindex/')
def encode_subtype(subtype):
seq_file = '../data/' + subtype + '_seqs'
with open('./aaindex/indexKeyList_2_pickle', 'rb') as f:
idxList = pickle.load(f)
batch_idxList_encode(seq_file, idxList)