def build_exact(footprint):
"""Build a PSSM based on p.footprint"""
PSSM = []
for char in footprint.lower():
PSSM.append(exact_codes[char])
return PSSM
def build_exact(footprint): """Build a PSSM based on p.footprint""" PSSM = [] for char in footprint.lower(): PSSM.append(exact_codes[char]) return PSSM
def distance(self, rhs, method='MSE'):
"""Calculate the distance between myself and another PPR using method:
method = 'MSE': Mean Squared Error
'KL': KL divergance
"""
if method == None:
method = "MSE"
return PSSM.distance(self.pssm, rhs.pssm, method)
def distance(self, rhs, method="MSE"):
"""Calculate the distance between myself and another PPR using method:
method = 'MSE': Mean Squared Error
'KL': KL divergance
"""
if method == None:
method = "MSE"
return PSSM.distance(self.pssm, rhs.pssm, method)
def nwprofile(S, gap, seq, residues,sc):
'''A function that performs the global alignment of a sequence to a profile'''
f=frequency(S,residues)
M=PSSM(seq,sc,residues,S)
F=[[0.0] * (len(f)+1) for i in range (0, len(seq)+1)]
for i in range(1, len(seq)+1):
for j in range(1, len(f)+1):
F[i][j]=max([F[i-1][j]+gap, F[i][j-1]+gap, F[i-1][j-1]+M[i-1][j-1]])
return F[i][j]
return F
def load_pprs():
pprs = list(SeqIO.parse(os.path.join(utils.TestDir, "Known_PPRs.gb"), "gb",
alphabet=Alphabet.generic_dna))
#apply known binding sites
f = open(os.path.join(utils.TestDir, "known_sites.json"))
sites = json.load(f)
for name,location in sites.iteritems():
for p in pprs:
if p.name.lower() == name.lower():
p.footprints = [l.replace('u','t') for l in location]
#build PSSMs
for p in pprs:
p.barkan = PSSM.build(p, 'barkan')
p.yagi = PSSM.build(p, 'yagi')
p.exact = [build_exact(footprint) for footprint in p.footprints]
return pprs
def load_pprs():
pprs = list(
SeqIO.parse(os.path.join(utils.TestDir, "Known_PPRs.gb"),
"gb",
alphabet=Alphabet.generic_dna))
#apply known binding sites
f = open(os.path.join(utils.TestDir, "known_sites.json"))
sites = json.load(f)
for name, location in sites.iteritems():
for p in pprs:
if p.name.lower() == name.lower():
p.footprints = [l.replace('u', 't') for l in location]
#build PSSMs
for p in pprs:
p.barkan = PSSM.build(p, 'barkan')
p.yagi = PSSM.build(p, 'yagi')
p.exact = [build_exact(footprint) for footprint in p.footprints]
return pprs
def PSSM_single_test(length=1000):
seq = Seq(''.join(random.choice('ACGT') for x in range(length-len(target))))
pos = random.randrange(0, len(seq))
seq = seq[0:pos] + str(target.seq) + seq[pos:]
aln = PSSM.search(ppr,seq,gaps=2)
rank = -1
for i,a in enumerate(aln):
if a.pos == pos+1:
rank = i
break
return rank
def PSSM_single_test(length=1000):
seq = Seq(''.join(
random.choice('ACGT') for x in range(length - len(target))))
pos = random.randrange(0, len(seq))
seq = seq[0:pos] + str(target.seq) + seq[pos:]
aln = PSSM.search(ppr, seq, gaps=2)
rank = -1
for i, a in enumerate(aln):
if a.pos == pos + 1:
rank = i
break
return rank
def smprofile(S, gap, seq, residues,sc):
'''A function that performs the local alignment of a sequence to a profile'''
f=frequency(S,residues)
M=PSSM(seq,sc,residues,S)
F=[[0.0] * (len(f)+1) for i in range (0, len(seq)+1)]
for i in range(1, len(seq)+1):
for j in range(1, len(f)+1):
F[i][j]=max([F[i-1][j]+gap, F[i][j-1]+gap, F[i-1][j-1]+M[i-1][j-1],0.0])
maxscore=0
for i in range(1, len(seq)+1):
for j in range(1, len(f)+1):
if F[i][j]>maxscore:
maxscore=F[i][j]
print F
print maxscore
def get_domains(ppr, plastid, percentile=10.0, gaps=1, type='PPR'): """Get a list of putative binding domains in the plastid""" feats = PSSM.search(ppr,plastid,gaps=gaps,show_stats=False) if feats: top = feats[0].qualifiers['odds'] bottom = feats[-1].qualifiers['odds'] threshold = top - (percentile/100.0) * (top-bottom) for f in feats: f.type = type for i in range(len(feats)): if feats[i].qualifiers['odds'] < threshold: return feats[0:i-1] return feats
def __init__(self, seq_record):
self.seq_record = seq_record
self.pssm = PSSM.build(self.seq_record)
def __init__(self, seq_record):
self.seq_record = seq_record
self.pssm = PSSM.build(self.seq_record)