def f_read_tffm(tffm_dir, tf_name):
import tffm_module
from constants import TFFM_KIND
file_list = os.listdir(tffm_dir)
if tf_name.lower() != "ctcf" and tf_name.lower() != "ebf1":
print "TFFM not found for %s" % tf_name
print "[Warning: change tf to CTCF motif]"
tf_name = "ebf1"
pattern = "%s.*detail.*xml" % (tf_name)
matched_file = grep_list(pattern, file_list)
if len(matched_file) == 1:
tffm_file = tffm_dir + grep_list(pattern, file_list)[0]
#logging.debug(tffm_file)
#tffm_first_order = tffm_module.tffm_from_xml(tffm_file, TFFM_KIND.FIRST_ORDER)
tffm_detailed = tffm_module.tffm_from_xml(tffm_file,
TFFM_KIND.DETAILED)
return tffm_detailed
else:
return None
def find_tffm_hits(xml, seq_file):
""" Predict hits in sequences using a TFFM. """
#import sys
#sys.path.append("/raid6/amathelier/TFFM+DNAshape/bin/TFFM/")
import tffm_module
from constants import TFFM_KIND # TFFM-framework
from hit_module import HIT
tffm = tffm_module.tffm_from_xml(xml, TFFM_KIND.FIRST_ORDER)
return [
hit for hit in tffm.scan_sequences(seq_file, only_best=True) if hit
]
def find_tffm_hits(xml, seq_file, tffm_kind):
""" Predict hits in sequences using a TFFM. """
#import sys
#sys.path.append("/raid6/amathelier/TFFM+DNAshape/bin/TFFM/")
import tffm_module
from constants import TFFM_KIND # TFFM-framework
if tffm_kind == 'first_order':
tffm_kind = TFFM_KIND.FIRST_ORDER
elif tffm_kind == 'detailed':
tffm_kind = TFFM_KIND.DETAILED
else:
sys.exit('The type of TFFM should be "first_order" or "detailed".')
tffm = tffm_module.tffm_from_xml(xml, tffm_kind)
return [hit for hit in
tffm.scan_sequences(seq_file, only_best=True) if hit]
Mdata = num.findall(matrix)
#print("list(reversed(Mdata)) : ",list(reversed(Mdata)))
matScore, lenMotif = get_score_matrix(Mdata, matrixType, pseudoCount)
# The following line allows to produce the reversed matrix
'''if we take the example given before : A T G C
Position 1: 0.4444 0.155 0.654 0.645
Position 2: 0.1645 0.1565 0.21614 0.16456
Now, we can notice that scores change between the positions 1 and 2, and between A and T, and between G and C.
So we can calculate with this reverse matrix, the score of the complementary strand.
'''
matRev = list(reversed(matScore))
tffm_first_order = ""
else:
matRev = ""
matScore = ""
tffm_first_order = tffm_module.tffm_from_xml(tffm, TFFM_KIND.FIRST_ORDER)
lenMotif = tffm_first_order.__len__() + 2
########## get INTERDISTANCE VALUES for POSITIVE sets:
if not pos_flag:
len_pos = 0
sequence_number = 0
with open(FastaFile, "r") as f1:
for line in f1:
if line.find(">") != -1:
line = line.strip()
line = line.replace("-", ":")
line = line.split(":")
len_pos += float(line[2]) - float(line[1]) + 1 - lenMotif
sequence_number += 1
out = open("tffm_first_order_dense_logo.svg", "w")
tffm_first_order.print_dense_logo(out)
out.close()
tffm_detailed = tffm_module.tffm_from_meme("meme.txt", TFFM_KIND.DETAILED)
tffm_detailed.write("tffm_detailed_initial.xml")
tffm_detailed.train("train.fa")
tffm_detailed.write("tffm_detailed.xml")
out = open("tffm_detailed_summary_logo.svg", "w")
tffm_detailed.print_summary_logo(out)
out.close()
out = open("tffm_detailed_dense_logo.svg", "w")
tffm_detailed.print_dense_logo(out)
out.close()
tffm_first_order = tffm_module.tffm_from_xml("tffm_first_order.xml",
TFFM_KIND.FIRST_ORDER)
print "1st-order all"
for hit in tffm_first_order.scan_sequences("test.fa"):
if hit:
print hit
print "1st-order best"
for hit in tffm_first_order.scan_sequences("test.fa", only_best=True):
print hit
tffm_detailed = tffm_module.tffm_from_xml("tffm_detailed.xml",
TFFM_KIND.DETAILED)
print "detailed all"
for hit in tffm_detailed.scan_sequences("test.fa"):
if hit:
print hit