def score_sequence_set(sequence_set, pssms, p_binding_site):
models = [
build_hmm_from_semi_parsed(
parsed,
p_binding_site=p_binding_site
)
for parsed
in pssms
]
return calculate_model_counts_on_sequences(models, sequence_set)
def get_roc_for_sequences(p_binding_site, positive_sequences, negative_sequences, pssms):
models = [
build_hmm_from_semi_parsed(
parsed,
p_binding_site=p_binding_site
)
for parsed
in pssms
]
roc = RocCalculator()
update_roc(roc, generate_roc_data(models, positive_sequences, negative_sequences))
return roc
def score_sequence_set(sequence_set, pssms, p_binding_site):
models = [
build_hmm_from_semi_parsed(parsed, p_binding_site=p_binding_site)
for parsed in pssms
]
return calculate_model_counts_on_sequences(models, sequence_set)
for fragment, pssm in pssms():
sequence_file = os.path.join(sequence_dir,
sequence_filename_fmt % fragment)
model_file = os.path.join(model_dir, '%s-%s.pssm' % (fragment, pssm))
logging.info('Loading sequences: %s', sequence_file)
sequences = list(sequences_from_fasta(sequence_file))
numpy_seqs = map(seq_to_numpy, sequences)
logging.info('Loaded %d sequences', len(sequences))
logging.info('Parsing PSSMs: %s', model_file)
pssms = list(parse_models(open(model_file)))
logging.info('Building models')
models = [
build_hmm_from_semi_parsed(parsed, p_binding_site=p_binding_site)
for parsed in pssms
]
def nucleotide_dist():
return numpy.zeros(4) + .25
base_dists = DictOf(nucleotide_dist)
min_site_length = 20
logging.info('Analysing sequences')
for hmm, traits in models:
sites = []
for sequence in numpy_seqs:
# analyse the sequence for its most likely state sequence
sequence_file = os.path.join(sequence_dir, sequence_filename_fmt % fragment)
model_file = os.path.join(model_dir, '%s-%s.pssm' % (fragment, pssm))
logging.info('Loading sequences: %s', sequence_file)
sequences = list(sequences_from_fasta(sequence_file))
numpy_seqs = map(seq_to_numpy, sequences)
logging.info('Loaded %d sequences', len(sequences))
logging.info('Parsing PSSMs: %s', model_file)
pssms = list(parse_models(open(model_file)))
logging.info('Building models')
models = [
build_hmm_from_semi_parsed(parsed, p_binding_site=p_binding_site)
for parsed in pssms
]
def nucleotide_dist():
return numpy.zeros(4) + .25
base_dists = DictOf(nucleotide_dist)
min_site_length = 20
logging.info('Analysing sequences')
for hmm, traits in models:
sites = []
for sequence in numpy_seqs:
# analyse the sequence for its most likely state sequence
LL, states = hmm.viterbi(sequence)
logging.info('Loading sequences: %s', options.sequences_file)
sequences = dict(sequences_from_fasta(options.sequences_file))
numpy_seqs = dict((desc, seq_to_numpy(seq)) for desc, seq in sequences.iteritems())
logging.info('Loaded %d sequences', len(sequences))
logging.info('Parsing PSSMs: %s', options.models_file)
pssms = list(parse_models(open(options.models_file)))
logging.info('Building models')
models = [
build_hmm_from_semi_parsed(parsed, p_binding_site=options.p_binding_site)
for parsed in pssms
]
logging.info('Analysing sequences')
p_binding_sites = list()
sites_file = open('sites.txt', 'w')
for hmm, traits in models:
for desc, sequence in numpy_seqs.iteritems():
p_binding_site, site_seq = analyse_sequence_for_best_site(sequence, hmm, traits)
p_binding_sites.append(p_binding_site)
site = numpy_to_seq(site_seq)
logging.info('%s: p(binding site)=%12g, sequence=%s', desc, p_binding_site, site)
print >> sites_file, '%s, %12g, %s' % (desc, p_binding_site, site)
sites_file.close()
