def __init__(self, pastafile, reffile, seqfile, verbose=False):
"""
Constructor
@param pastafile: filename for residue list
@type pastafile: string
@param reffile: filename for reference list (eg. BMRB ASCII file)
@type reffile: string
@param seqfile: filename for FASTA sequence file
@type seqfile: string
"""
fh = FileHandler()
## list of Residue.PastaResidue objects
self.residues = fh.read_pasta(pastafile, reffile)
## list of Residue.AminoAcid objects
self.amino_acids = fh.read_preset(reffile)
## list of Residue.AminoAcid objects
self.seq = fh.read_seq(seqfile, reffile)
self.P = None ## numpy.ndarray for typing posterior probabilities
self.L = None ## numpy.ndarray for linking constraints
self.S = None ## numpy.ndarray for aa type in sequence
self.A = None ## list of assignments and respective similarity score
## ILP STUFF
self.B = None ## list of assignments and respective costs
self.C = None ## cost matrix ILP
self.Xs = None ## assignment matrices from solution pool,
self.ILP_L = None ## Linking Matrix of ILP
self.typing_elapsed = 0
self.linking_elapsed = 0
self.mapping_elapsed = 0
self.full_running_time = 0
if __name__ == '__main__':
from FileHandler import FileHandler
from Linking import Linking
import pylab as pl
fh = FileHandler()
statsfile = 'shiftPresets/bmrb.shift'
folder = "Datasets/Ubiquitin/"
fn = folder + "residue_lists/Ub_opt_relabeled.pasta"
seqfile = folder + "sequences/Ub.fasta"
result_folder = folder + "Results/"
tol = .6
strat = "Joint"
residues = fh.read_pasta(fn, statsfile)
link = Linking(residues)
L = link.linking_matrix(strat, tolerance=tol, conservative=True)
L2 = link.linking_matrix(strat, tolerance=tol, conservative=False)
pl.matshow(L)
pl.matshow(L2)
pl.show()
## link.check_pos_constraints(L,residues,tol)
## link.check_neg_constraints(L,residues,tol)
#
# A = fh.assignment_matrix_from_assigned(fn, seqfile, statsfile).astype("i")
if __name__ == "__main__":
import pylab as pl
from numpy import max, mean, take, zeros
from Residue import PastaResidue, AminoAcid
from MaxLikelihood import Likelihood
from FileHandler import FileHandler
fh = FileHandler()
fn = "/is/ei/jhooge/EclipseWorkspaces/PASTA/PyPASTA/GMM/src/"\
"Classification/tests/reference_lists/bmrb.shift"
# pastalist = 'tests/residue_lists/all_singles.pasta'
pastalist = "Datasets/Ubiquitin/residue_lists/"\
"incompleteData/Ub_bmrb_missing_shifts_0.50.pasta"
statsfile = fn
amino_acids = fh.read_preset(fn)
residues = fh.read_pasta(pastalist, statsfile)
toAA = AbstractMapping.create("on_amino_acid")
toRes = AbstractMapping.create("on_residue")
aas = amino_acids
del aas[1]
L = Likelihood()
A = L.calc_likelihoods(residues, amino_acids, previous=False)
pl.matshow(A)
pl.colorbar()
pl.show()
x[i] = ones_like(x[i])
i += 1
return x
if __name__ == '__main__':
from FileHandler import FileHandler
from numpy import argmax, mean, max
import pylab as pl
fh = FileHandler()
L = Likelihood()
# pastafile = 'multiple_test_files/Ubiquitin/Ub_new.pasta'
pastafile = 'tests/residue_lists/all_singles.pasta'
statsfile = 'shiftPresets/bmrb.shift'
residues = fh.read_pasta(pastafile, statsfile)
amino_acids = fh.read_preset(statsfile)
print ' '.join([aa.three_let for aa in amino_acids])
# r = residues[1]
# print r.name
# p11 = L.calc_likelihoods(residues, amino_acids, previous=False, summarize=mean)
# p12 = L.calc_likelihoods(residues, amino_acids, previous=True, summarize=mean)
p21 = L.calc_likelihoods(residues, amino_acids, previous=False, summarize=max)
p22 = L.calc_likelihoods(residues, amino_acids, previous=True, summarize=max)
print p21
print p22
# pl.matshow(p11)
# pl.colorbar()
from FileHandler import FileHandler
from Mapping import Mapping
from Mapping2 import AbstractMapping
from Definitions import three2One
from MaxLikelihood import Likelihood
from numpy import array,mean, max
import pylab as pl
fh = FileHandler()
pairs = 'tests/residue_lists/all_pairs.pasta'
singles = 'tests/residue_lists/all_singles.pasta'
seqfile_pairs = 'tests/sequences/all_pairs.fasta'
seqfile_singles = 'tests/sequences/all_singles.fasta'
statsfile = 'tests/reference_lists/bmrb.shift'
single_residues = fh.read_pasta(singles, statsfile)
amino_acids = fh.read_seq(seqfile_singles, statsfile)
atoms = {'CO' :0, 'CA' :1,
'CB' :2, 'CG' :3,
'CG1':4, 'CG2':5,
'CD' :6, 'CD1':7,
'CD2':8, 'CE' :9,
'CE1':10, 'CE2':11,
'CE3':12, 'CZ':13, 'CZ2':14,
'CZ3':15, "CH2":16}
aa_names = ''.join([aa.one_let for aa in amino_acids])
# k = aa_names.index("F")
# print "Residue: ",res.name_im1, res.get_carbons(previous=True)
# print "Amino Acid: ",aa.three_let, aa.get_carbons()
mapper1 = Mapping(aa_mapping=True)
res= get_rnd_residue(residues,i)
swap_rnd_atom(res,previous)
i = get_indices(residues, previous)
m = float(get_no_of_ambiguous_keys(residues))
return residues
if __name__ == '__main__':
fh = FileHandler()
folder = "Datasets/Ubiquitin/"
pastafn = "residue_lists/Ub_bmrb_unassigned.pasta"
pastafn2 = "residue_lists/Ub_opt_unambiguous_unassigned.pasta"
presetfn = 'shiftPresets/bmrb.shift'
seqfn = folder + "sequences/Ub_bmrb.fasta"
result_folder = folder + "Results/"
residues = fh.read_pasta(folder + pastafn, presetfn)
residues2 = fh.read_pasta(folder + pastafn2, presetfn)
## ADD NOISE
# noise = linspace(0,3,31)
# for n in noise:
# print residues[0].shifts_i.values()
# new_res = add_noise(deepcopy(residues), n)
# outfolder = folder + "UbqBaxNoise=%.1f.pasta" %n
# fh.write_pasta(outfolder, new_res)
# print outfolder, "written"
## INTRODUCE AMBIGUOUS SHITS
# old_res = deepcopy(residues)
# for p in arange(.0, .55, .05):
# new_res = make_ambiguous(old_res,p)
