def score(self, test_case):
"Score the test case and return the hits."
record, seq, pssms = test_case
hits = B.score_pssms_on_sequence(make_sequence_vec(pssms),
str(seq.seq),
threshold=self.min_threshold)
return hits
def test_pssm_distributions():
pssm_acc = 'M00750'
transfac_pssms = biopsy.get_transfac_pssm_accessions(
biopsy.get_default_transfac_pssm_filter())
print 'Got', len(transfac_pssms), 'pssms'
# initialise the distributions
score_dists = {}
for tp in transfac_pssms:
score_dists[tp] = {}
for k in range(1, 100):
score_dists[tp][k] = 0
score_dists['all'] = {}
for k in range(1, 100):
score_dists['all'][k] = 0
# parse the mouse chromosome, score the pssms and fill in the distributions
bases = 0
seq = ''
start = time.clock()
for line in open(
'C:/Data/ensembl/chromosomes/Mus_musculus.NCBIM34.dec.dna.chromosome.1.fa',
'r'):
if line.startswith('>'): continue
seq += line.strip('\r\n').replace('N', '')
# Take 1kb at a time
if len(seq) >= 1000:
hits = biopsy.score_pssms_on_sequence(transfac_pssms, seq, 0.0)
for h in hits:
bin = int(100.0 * h.p_binding)
if 0 != bin:
# print h.binder, bin
score_dists[h.binder][bin] += 1
score_dists['all'][bin] += 1
bases += len(seq)
seq = ''
print 'Bases:', bases
if bases >= 2800000: break
elapsed = time.clock() - start
print 'Scored', len(
transfac_pssms), 'pssms on', bases, 'bases in', elapsed, 'seconds'
print 'Estimate for mouse chromosome 1 (secs):', elapsed * 190000000 / bases
print 'Estimate for mouse chromosome 1 (days):', elapsed * 190000000 / bases / 60 / 60 / 24
print 'Estimate for # bases/hour:', bases * 3600 / elapsed
# remember scores for later
f = open('pssm_p_binding_dists.txt', 'w')
print 'Writing pssm p(binding) distributions to:', f
pickle.dump(score_dists, f)
f.close()
def test_score_pssms():
# 'V$AP1_Q2'
transfac_pssms = biopsy.get_transfac_pssm_accessions(
biopsy.get_default_transfac_pssm_filter())
print 'Got', len(transfac_pssms), 'pssms'
seq = 'tacatcatctgtctgcagtagtctaaccgaccccccccagttttagaagcagactgcatgcggacgggaccgcggatcgcgcggtgcgcctcagtgtacttccgaacgaatgagtcattaatagagcgctatatcgtaactgtctttgacgaagtataccgaaaccgtgcagccagacgtgatccgggcgttgtaaaggcgatcagcgccctaggagtaccatttttgccgtaggcttgcgtctcaaagaccagctggggcgtggtatcactcgtcagtacgatttctgccagatagatagcatagactgaaccttaggcccaatagggacacaattacccgagtgactgactggtctaaggggagtccccccttaaaacgttttacgtaatagcgggctccagaagcaaagcatcggtttgagccccagtactaaacgtttgagtgtttgctctcgtctgataggtaaaccgacaagagaaccaagctcaaggcgcggtaggtgcgccttgcgaactgttgatgccgtgagcgccaccatcccgtgcatcataggcagggagagaagaccacatggccttgcgaccgtatgagctgtttcagattaaatgccaacgggcatggtcggtgtccagcattttttgcagtcagctggtggtacacagtggggacaagaacgcctctggtagatgtcttctgaaggagtaactcatttcgttgaatcgaccttcccttgcgcttgaacgcggacctctagtctctctcgcagactggggtcgaaaatcaaggtagatatggaatgttccgcatgagggtagcgaccggatcgggcgtcaagtatatcctccctgctacgtccccctactagcctcagtccgcctcgaacctaggaagattggccacatcagcttggtggatgcctggtccatacttcagacccgagaatgttagacaggaccccatttggctcctttacgtacgatctatgtagacgcagtga'
# seq = 'acatcat'
# seq = 'gat'
# hits = biopsy.HitVec()
hits = biopsy.score_pssms_on_sequence(transfac_pssms, seq, 0.05)
print hits
print 'score_pssm_on_sequence: Got', len(hits), 'hits from', len(
seq), 'bases'
hits = biopsy.analyse(seq, 0.05)
# print hits
print 'analyse: Got', len(hits), 'hits from', len(seq), 'bases'
def test_pssm_distributions():
pssm_acc = 'M00750'
transfac_pssms = biopsy.get_transfac_pssm_accessions( biopsy.get_default_transfac_pssm_filter() )
print 'Got', len( transfac_pssms ), 'pssms'
# initialise the distributions
score_dists = { }
for tp in transfac_pssms:
score_dists[ tp ] = { }
for k in range(1,100):
score_dists[ tp ][ k ] = 0
score_dists[ 'all' ] = { }
for k in range(1,100):
score_dists[ 'all' ][ k ] = 0
# parse the mouse chromosome, score the pssms and fill in the distributions
bases = 0
seq = ''
start = time.clock()
for line in open( 'C:/Data/ensembl/chromosomes/Mus_musculus.NCBIM34.dec.dna.chromosome.1.fa', 'r' ):
if line.startswith( '>' ): continue
seq += line.strip( '\r\n' ).replace( 'N', '' )
# Take 1kb at a time
if len( seq ) >= 1000:
hits = biopsy.score_pssms_on_sequence( transfac_pssms, seq, 0.0 )
for h in hits:
bin = int( 100.0 * h.p_binding )
if 0 != bin:
# print h.binder, bin
score_dists[ h.binder ][ bin ] += 1
score_dists[ 'all' ][ bin ] += 1
bases += len( seq )
seq = ''
print 'Bases:', bases
if bases >= 2800000: break
elapsed = time.clock() - start
print 'Scored', len( transfac_pssms ), 'pssms on', bases, 'bases in', elapsed, 'seconds'
print 'Estimate for mouse chromosome 1 (secs):', elapsed * 190000000 / bases
print 'Estimate for mouse chromosome 1 (days):', elapsed * 190000000 / bases / 60 / 60 / 24
print 'Estimate for # bases/hour:', bases * 3600 / elapsed
# remember scores for later
f = open( 'pssm_p_binding_dists.txt', 'w' )
print 'Writing pssm p(binding) distributions to:', f
pickle.dump(score_dists, f)
f.close()
def test_score_pssms():
# 'V$AP1_Q2'
transfac_pssms = biopsy.get_transfac_pssm_accessions( biopsy.get_default_transfac_pssm_filter() )
print 'Got', len( transfac_pssms ), 'pssms'
seq = 'tacatcatctgtctgcagtagtctaaccgaccccccccagttttagaagcagactgcatgcggacgggaccgcggatcgcgcggtgcgcctcagtgtacttccgaacgaatgagtcattaatagagcgctatatcgtaactgtctttgacgaagtataccgaaaccgtgcagccagacgtgatccgggcgttgtaaaggcgatcagcgccctaggagtaccatttttgccgtaggcttgcgtctcaaagaccagctggggcgtggtatcactcgtcagtacgatttctgccagatagatagcatagactgaaccttaggcccaatagggacacaattacccgagtgactgactggtctaaggggagtccccccttaaaacgttttacgtaatagcgggctccagaagcaaagcatcggtttgagccccagtactaaacgtttgagtgtttgctctcgtctgataggtaaaccgacaagagaaccaagctcaaggcgcggtaggtgcgccttgcgaactgttgatgccgtgagcgccaccatcccgtgcatcataggcagggagagaagaccacatggccttgcgaccgtatgagctgtttcagattaaatgccaacgggcatggtcggtgtccagcattttttgcagtcagctggtggtacacagtggggacaagaacgcctctggtagatgtcttctgaaggagtaactcatttcgttgaatcgaccttcccttgcgcttgaacgcggacctctagtctctctcgcagactggggtcgaaaatcaaggtagatatggaatgttccgcatgagggtagcgaccggatcgggcgtcaagtatatcctccctgctacgtccccctactagcctcagtccgcctcgaacctaggaagattggccacatcagcttggtggatgcctggtccatacttcagacccgagaatgttagacaggaccccatttggctcctttacgtacgatctatgtagacgcagtga'
# seq = 'acatcat'
# seq = 'gat'
# hits = biopsy.HitVec()
hits = biopsy.score_pssms_on_sequence(
transfac_pssms,
seq,
0.05 )
print hits
print 'score_pssm_on_sequence: Got', len( hits ), 'hits from', len( seq ), 'bases'
hits = biopsy.analyse(
seq,
0.05)
# print hits
print 'analyse: Got', len( hits ), 'hits from', len( seq ), 'bases'
def __call__(self, sequence):
import biopsy
import biopsy.analyse_remos.consolidate_hits as CH
hits = biopsy.score_pssms_on_sequence(self.pssms, sequence.strip('Nn'), self.threshold)
return CH.maximal_chain_hits(hits)
def score(self, test_case):
"Score the test case and return the hits."
record, seq, pssms = test_case
hits = B.score_pssms_on_sequence(make_sequence_vec(pssms), str(seq.seq), threshold=self.min_threshold)
return hits
def score_pssms_on_record(pssms, record):
import biopsy
promoter_hits = biopsy.score_pssms_on_sequence(make_sequence_vec(pssms), record.seq.tostring())
promoter_p_binding = calculate_p_binding_from_hits(promoter_hits)
return promoter_p_binding
