logging.basicConfig(level=logging.INFO)
max_mosaics = 15
max_order = 5
me = evaluate_mosaics(max_mosaics=max_mosaics, max_order=max_order)
for i in range(max_order):
e = me[i * max_mosaics:(i + 1) * max_mosaics]
plot([x[1] for x in e], [x[2] for x in e])
xlabel('# mosaics')
ylabel('LL')
title('Evaluation of mosaic models of various Markov orders')
savefig('mosaic-evaluation.png', format='PNG')
raise
# load our sequences
sequences = convert_fasta_sequences(fasta_file_for_fragment('T00671'))
# build our model
model_by_states = create_mosaic_model(num_mosaics=1,
p_transition=0.,
alphabet_size=4,
order=2,
dirichlet_prior_strength=10.)
model = hmm.as_model(model_by_states)
print model.B
# convert our sequences to the correct order
sequences_order_n = [model.converter.to_order_n(s) for s in sequences]
#from IPython.Debugger import Pdb; Pdb().set_trace()
def callback(LL):
logging.basicConfig(level=logging.INFO)
max_mosaics = 15
max_order = 5
me = evaluate_mosaics(max_mosaics=max_mosaics, max_order=max_order)
for i in range(max_order):
e=me[i*max_mosaics:(i+1)*max_mosaics]
plot([x[1] for x in e], [x[2] for x in e])
xlabel('# mosaics')
ylabel('LL')
title('Evaluation of mosaic models of various Markov orders')
savefig('mosaic-evaluation.png', format='PNG')
raise
# load our sequences
sequences = convert_fasta_sequences(fasta_file_for_fragment('T00671'))
# build our model
model_by_states = create_mosaic_model(
num_mosaics=1,
p_transition=0.,
alphabet_size=4,
order=2,
dirichlet_prior_strength=10.
)
model = hmm.as_model(model_by_states)
print model.B
# convert our sequences to the correct order
sequences_order_n = [model.converter.to_order_n(s) for s in sequences]
print reduce_sequence(N.arange(10), 10)
print reduce_sequence(N.arange(10), 9)
print reduce_sequence(N.arange(10), 6)
print reduce_sequence(N.arange(10), 5)
print reduce_sequence(N.arange(10), 3)
print reduce_sequence(N.arange(10), 2)
print reduce_sequence(N.arange(10), 1)
raise RuntimeError('Stopping')
K = 8
model_file = 'bg-model.pickle'
fasta = '/home/john/Data/GappedPssms/apr-2009/T00759trimRM.fa'
bg_model = cPickle.load(open(model_file))
sequences = convert_fasta_sequences(fasta)
converted_seqs = [bg_model.converter.to_order_n(s) for s in sequences]
def k_mer_log_likelihoods_new(alpha, c):
result = N.empty(len(c)-K+1)
alpha_sum = alpha.sum(axis=1)
for i in xrange(len(c)-K+1):
if 0 == i:
result[i] = alpha_sum[K] / c[:K+1].prod()
else:
result[i] = alpha_sum[i+K-1] / alpha_sum[i-1] / c[i:i+K].prod()
return N.log(result)
def calculate_k_mer_scores(bg_model, converted_seqs, K):
for option in option_parser.option_list:
if option.dest:
logging.info('%s: %s (%s)', option.dest,
str(getattr(options, option.dest)), option.help)
#inputs = [('K10-g0.50-N200-L200-seed4-1', fasta_file_for_synthetic_data('K10-g0.50-N200-L200-seed4-1'))]
inputs = [(fragment, fasta_file_for_fragment(fragment))
for fragment in test_set_fragments]
# for each input sequence
for seq_tag, fasta_file in inputs:
# add a file handler to log for this test set
file_handler = logging.FileHandler(
os.path.join(options.output_dir, '%s.log' % seq_tag))
logging.getLogger('').addHandler(file_handler)
try:
sequences = convert_fasta_sequences(fasta_file)
#sequences = [s[:200] for s in sequences[:10]]
# set up the options for this test set
options.tag = seq_tag
options.bg_model_filename = "%s-bg-model.pickle" % seq_tag
# Run the algorithm
algorithm = SingleGapAlgorithm(options)
algorithm(sequences)
finally:
logging.getLogger('').removeHandler(file_handler)
file_handler.close()
"-f",
"--fasta",
dest="fasta",
help="The fasta file containing the sequences to run.")
add_algorithm_options(option_parser)
options, args = option_parser.parse_args()
if not options.output_dir:
raise ValueError('No output directory specified')
if not os.path.exists(options.output_dir):
os.makedirs(options.output_dir)
log_filename = os.path.join(options.output_dir, '%s.log' % options.tag)
file_handler = logging.FileHandler(log_filename)
file_handler.setFormatter(
logging.Formatter("%(asctime)s - %(levelname)s - %(message)s"))
logging.getLogger('').addHandler(file_handler)
logging.info('Writing log to %s', log_filename)
logging.info('Command line: %s', ' '.join(sys.argv))
for option in option_parser.option_list:
if option.dest:
logging.info('%32s: %s', option.dest,
str(getattr(options, option.dest)))
logging.info('Reading sequences from: %s' % options.fasta)
sequences = convert_fasta_sequences(options.fasta)
#sequences = [s[:200] for s in sequences[:10]]
# Run the algorithm
algorithm = SingleGapAlgorithm(options)
algorithm(sequences)
"-f",
"--fasta",
dest="fasta",
help="The fasta file containing the sequences to run."
)
add_algorithm_options(option_parser)
options, args = option_parser.parse_args()
if not options.output_dir:
raise ValueError('No output directory specified')
if not os.path.exists(options.output_dir):
os.makedirs(options.output_dir)
log_filename = os.path.join(options.output_dir, '%s.log' % options.tag)
file_handler = logging.FileHandler(log_filename)
file_handler.setFormatter(logging.Formatter("%(asctime)s - %(levelname)s - %(message)s"))
logging.getLogger('').addHandler(file_handler)
logging.info('Writing log to %s', log_filename)
logging.info('Command line: %s', ' '.join(sys.argv))
for option in option_parser.option_list:
if option.dest:
logging.info('%32s: %s', option.dest, str(getattr(options, option.dest)))
logging.info('Reading sequences from: %s' % options.fasta)
sequences = convert_fasta_sequences(options.fasta)
#sequences = [s[:200] for s in sequences[:10]]
# Run the algorithm
algorithm = SingleGapAlgorithm(options)
algorithm(sequences)
logging.info('%32s: %s', option.dest, str(getattr(options, option.dest)))
#
# For each data set
#
for fragment in test_set_fragments:
for i in xrange(1, cross_folds+1):
train_fasta, test_fasta = fasta_files_for_fragment_cross_fold(fragment, i)
logging.info('Training data set: %s', train_fasta)
#log.info('Test data set: %s', test_fasta)
seq_tag = '%s-%d' % (fragment, i)
file_handler = logging.FileHandler(os.path.join(options.output_dir, '%s.log' % seq_tag))
logging.getLogger('').addHandler(file_handler)
try:
sequences = convert_fasta_sequences(train_fasta)
#sequences = [s[:200] for s in sequences[:100]]
# set up the options for this test set
options.tag = seq_tag
options.bg_model_filename = "%s-bg-model.pickle" % seq_tag
# Run the algorithm
algorithm = SingleGapAlgorithm(options)
algorithm(sequences)
finally:
logging.getLogger('').removeHandler(file_handler)
file_handler.close()
