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__": 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()
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() # pl.matshow(p21)