def block2pssm(self, block_data, seq):
pssm_info = []
for i in range(len(block_data)):
score_dict = {}
for a in self.alphabet:
score_dict[a] = block_data[i][self.order[a]]
pssm_info.append((seq[i], score_dict))
return PSSM(pssm_info)
def from_sequences(
cls,
sequences,
name="unnamed",
pseudocounts="jaspar",
threshold=None,
relative_threshold=None,
):
"""Return a PSSM pattern computed from same-length sequences.
Parameters
----------
sequences
A list of same-length sequences
name
Name to give to the pattern (will appear in reports etc.)
pseudocounts
Either a dict {"A": 0.01, "T": ...} or "jaspar" for automatic
pseudocounts from the Biopython.motifs.jaspar module (recommended),
or None for no pseudocounts at all (not recommended!)
threshold
locations of the sequence with a PSSM score above this value will be
considered matches. For convenience, a relative_threshold can be
given instead.
relative_threshold
Value between 0 and 1 from which the threshold will be auto-computed.
0 means "match everything", 1 means "only match the one (or several)
sequence(s) with the absolute highest possible score".
"""
sequences = [Seq(s) for s in sequences]
motif = motifs.create(sequences)
cls.apply_pseudocounts(motif, pseudocounts)
pssm = PSSM(motif.pssm)
pssm.name = name
return MotifPssmPattern(
pssm=pssm,
threshold=threshold,
relative_threshold=relative_threshold,
)
def parse_pssm(fname):
pssm = []
with open(fname) as f:
for line in f.readlines():
token = line.rstrip('\r\n').split()
if len(token) == 0:
continue
if re.match(r'\d+', token[0]):
pssm.append(
(token[1],
dict((x, int(y)) for x, y in zip(AA, token[2:22]))))
return PSSM(pssm)