def weblogoPOIM(logofile, poim, max_len):
"""instead of plotting the POIM heatmap, create a weblogo from the 1st-degree poim"""
warnings.filterwarnings('ignore', ' This call to matplotlib.use()*')
from corebio.data import rna_letters, dna_letters, amino_acid_letters
from weblogolib import LogoData, LogoOptions, LogoFormat, classic, png_print_formatter
#print "WEBLOGO!"
#print "Writing ", logofile
#print poim[0]
positive_logo = []
negative_logo = []
for i in xrange(len(poim[0])):
positive_logo.append([])
negative_logo.append([])
for j in xrange(len(poim[0][i])):
if poim[0][i][j] < 0:
positive_logo[i].append(0)
negative_logo[i].append(poim[0][i][j] * -10000)
else:
negative_logo[i].append(0)
positive_logo[i].append(poim[0][i][j] * 1000)
#print "Positive logo: ", positive_logo
#print "Negative logo: ", negative_logo
pos_data = LogoData.from_counts('ACGT', numpy.array(positive_logo).T, None)
neg_data = LogoData.from_counts("ACGT", numpy.array(negative_logo).T, None)
neg_opt = LogoOptions()
neg_opt.fineprint += " from KIRMES POIM data"
#logoopt.number_interval = 5
neg_opt.small_fontsize = 4
neg_opt.title_fontsize = 8
neg_opt.scale_width = False
title = os.path.split(logofile)[1]
title = title[:title.rfind(".")]
if "_" in title:
title = title[title.rfind("_") + 1:]
neg_opt.logo_title = title + " Negative Logo"
neg_format = LogoFormat(neg_data, neg_opt)
pos_opt = LogoOptions()
#pos_opt.show_ends = True
pos_opt.scale_width = False
pos_opt.logo_title = title + " Positive Sequence Logo"
pos_opt.show_fineprint = False
pos_opt.color_scheme = classic
pos_format = LogoFormat(pos_data, pos_opt)
neg_logo = open(logofile + "n.png", 'w')
png_print_formatter(neg_data, neg_format, neg_logo)
neg_logo.close()
pos_logo = open(logofile + "p.png", 'w')
png_print_formatter(pos_data, pos_format, pos_logo)
pos_logo.close()
concatPNG(logofile, (logofile + "p.png", logofile + "n.png"))
os.remove(logofile + "n.png")
os.remove(logofile + "p.png")
def create_logo(self, seqs=[]):
"""Create sequence logo for input sequences."""
# seperate headers
headers, instances = [list(x)
for x in zip(*seqs)]
if self.options.sequence_type is 'rna':
alphabet = Alphabet('ACGU')
elif self.options.sequence_type is 'protein':
alphabet = Alphabet('ACDEFGHIKLMNPQRSTVWY')
else:
alphabet = Alphabet('AGCT')
motif_corebio = SeqList(alist=instances, alphabet=alphabet)
data = wbl.LogoData().from_seqs(motif_corebio)
format = wbl.LogoFormat(data, self.options)
if self.output_format == 'png':
return wbl.png_formatter(data, format)
elif self.output_format == 'png_print':
return wbl.png_print_formatter(data, format)
elif self.output_format == 'jpeg':
return wbl.jpeg_formatter(data, format)
else:
return wbl.eps_formatter(data, format)
def generate_logos(motifs, foldername, filetype='png'):
options = wl.LogoOptions()
options.color_scheme = wl.std_color_schemes["chemistry"]
for i, motif in enumerate(motifs):
if motif:
my_format = wl.LogoFormat(motif.data, options)
if 'png' in filetype:
to_write = wl.png_print_formatter(motif.data, my_format)
foo = open(
os.path.join(foldername,
str(i) + '_' + str(len(motif.seqs)) + ".png"),
"wb")
foo.write(to_write)
foo.close()
if 'pdf' in filetype:
to_write = wl.pdf_formatter(motif.data, my_format)
foo = open(
os.path.join(foldername,
str(i) + '_' + str(len(motif.seqs)) + ".pdf"),
"wb")
foo.write(to_write)
foo.close()
if 'txt' in filetype:
to_write = ''.join(["%s\n" % str(seq) for seq in motif.seqs])
foo = open(
os.path.join(foldername,
str(i) + '_' + str(len(motif.seqs)) + ".txt"),
"w")
foo.write(to_write)
foo.close()
else:
raise ValueError(
'Invalid filetype. Available options: png, pdf or txt. ')
def create_logo(self, seqs=[]):
"""Create sequence logo for input sequences."""
# seperate headers
headers, instances = [list(x)
for x in zip(*seqs)]
if self.options.sequence_type is 'rna':
alphabet = Alphabet('ACGU')
elif self.options.sequence_type is 'protein':
alphabet = Alphabet('ACDEFGHIKLMNPQRSTVWY')
else:
alphabet = Alphabet('AGCT')
motif_corebio = SeqList(alist=instances, alphabet=alphabet)
data = wbl.LogoData().from_seqs(motif_corebio)
format = wbl.LogoFormat(data, self.options)
if self.output_format == 'png':
return wbl.png_formatter(data, format)
elif self.output_format == 'png_print':
return wbl.png_print_formatter(data, format)
elif self.output_format == 'jpeg':
return wbl.jpeg_formatter(data, format)
else:
return wbl.eps_formatter(data, format)
def create_logo(input_seqs_fname, logo_fname, options):
"""
Create a logo plot png using weblogo from a fasta created with write_logo_input()
"""
with open(input_seqs_fname, "rU") as f:
seqs = w.read_seq_data(f)
data = w.LogoData.from_seqs(seqs)
subprocess.check_call(("rm " + input_seqs_fname).split())
format = w.LogoFormat(data, options)
with open(logo_fname, "w") as f:
f.write(w.png_print_formatter(data, format))
def write_weblogo(self, filepath):
matrix_tuple = []
for distribution in self.values:
matrix_tuple.append(tuple(distribution))
dataArray = np.array(tuple(matrix_tuple))
alph = Alphabet(''.join(self.alphabet))
weblogoData = LogoData.from_counts(alph, dataArray)
weblogoOptions = LogoOptions(color_scheme=classic)
weblogoOptions.title = "PWM"
weblogoFormat = LogoFormat(weblogoData, weblogoOptions)
weblogo_file = open(filepath, 'w')
weblogo_file.write(png_print_formatter(weblogoData, weblogoFormat))
weblogo_file.close()
with open(args.output_base + ".fasta", "rU") as f:
seqs = w.read_seq_data(f)
data = w.LogoData.from_seqs(seqs)
subprocess.check_call(("rm " + args.output_base + ".fasta").split())
options = w.LogoOptions()
options.unit_name = "probability"
options.yaxis_label = "Probability"
options.xaxis_label = "Site Position"
options.show_fineprint = False
options.stacks_per_line = 500
options.tic_length = 10
format = w.LogoFormat(data, options)
with open(args.output_base + ".png", 'w') as f:
f.write(w.png_print_formatter(data, format))
aa_naive_seqs_c = Counter(aa_naive_seqs)
num_trees = len(aa_naive_seqs)
aa_naive_seqs_d = OrderedDict(
("naive_" + str(i) + "_" + str(float(count) / num_trees), seq)
for i, (seq, count) in enumerate(aa_naive_seqs_c.most_common(None))
)
write_to_fasta(aa_naive_seqs_d, args.output_base + ".fasta")
aa_dna_naive_seqs_d = {}
for k, g in groupby(naive_seqs, lambda seq: translate(seq)):
if k in aa_dna_naive_seqs_d:
aa_dna_naive_seqs_d[k].update(g)
else:
aa_dna_naive_seqs_d[k] = Counter(g)
