def write_groups(groups, out_csv):
headers = ['group', 'other_groups', 'align', 'consistent', 'sequence']
titles = groups[0]['peptides'][0]['attr'].keys()
headers.extend(titles)
rows = [headers]
for group in groups:
calc_alignment(group)
check_misalignment(group)
for peptide in group['peptides']:
row = []
row.append(group['i'])
other_groups = []
for i_group in peptide['groups']:
if i_group != group['i']:
other_groups.append(str(i_group))
if other_groups:
other_groups = ';'.join(other_groups)
else:
other_groups = ''
row.append(other_groups)
row.append(group['alignment'])
row.append(not group['misaligned'])
row.append(peptide['sequence'])
for title in titles:
if title in peptide['attr']:
row.append(peptide['attr'][title])
else:
row.append('')
row.append(' ')
for c in peptide['aligned_sequence']:
row.append(c)
rows.append(row)
rows.append([
group['i'],
'',
group['alignment'],
not group['misaligned'],
])
datafile.write_csv(rows, out_csv)
def run(
motif_csv, test_fasta, out_html, aa_probs,
variant='fixed',
alpha=1.0, beta=0.01):
log_odds_pssm = get_motif_log_odds(
motif_csv, aa_probs,
alpha=alpha, beta=beta,
variant=variant)
write_log_odds(log_odds_pssm, motif_csv.replace('.pssm', '').replace('.csv', '.pssm'))
n_position = len(log_odds_pssm.keys())
seqids, fastas = uniprot.read_fasta(test_fasta)
saved_scores = []
html = "<body>"
for seqid, entry in fastas.items():
# print "Sequence", seqid
test_seq = entry['sequence']
scores = score_full_sequence(log_odds_pssm, test_seq)
html += "<h1>%s</h1>" % entry['description']
html += make_html_for_seq_scores(test_seq, n_position, scores)
saved_scores.append({
'seqid': seqid,
'description': entry['description'],
'scores': scores,
'seq': test_seq
})
html += "</body>"
open(out_html, 'w').write(html)
titles = ['seqid', 'description', 'position', 'seq', 'score']
rows = [titles]
for entry in saved_scores:
n_seq = len(entry['seq'])
for i in range(0, n_seq - n_position):
score = entry['scores'][i]
pos = i + 1
if score > 0:
row = [
entry['seqid'],
entry['description'],
pos,
score,
entry['seq'][i:i+n_position]
]
rows.append(row)
datafile.write_csv(rows, out_html + '.csv')
def write_overlap_kernels(groups, out_csv):
headers = ['group', 'size_group', 'sequence']
rows = [headers]
for group in groups:
rows.append([
group['i'],
len(group['peptides']),
get_kernel(group)
])
datafile.write_csv(rows, out_csv)
def write_subset_kernels(groups, out_csv):
headers = ['group', 'size_group', 'alignment', 'sequence']
rows = [headers]
for group in groups:
rows.append([
group['i'],
len(group['peptides']),
group['alignment'],
group['shortest']['sequence'],
])
datafile.write_csv(rows, out_csv)
def make_csv(data, csv_fname):
rows = []
header = ['seqid', 'description' 'n_res', 'DP', 'DQ', 'DR', 'all']
rows.append(header)
for seqid in data:
protein = data[seqid]
n_res = protein['length']
sources = protein['sources']
n_source = len(sources)
row = [seqid, protein['description']]
coverage = [[0] * n_res for i in range(n_source + 1)]
for i_source in range(n_source):
source = sources[i_source]
for i, j in source['intervals']:
for k in range(i, j):
coverage[i_source][k] = 1
coverage[-1][k] = 1
row.append(sum(coverage[i_source]) / float(n_res) * 100.0)
row.append(sum(coverage[-1]) / float(n_res) * 100.0)
rows.append(row)
datafile.write_csv(rows, csv_fname)
left = -min(indices)
max_len = 0
for index, seq in zip(indices, sequences):
str_len = left + index + len(seq)
if str_len > max_len:
max_len = str_len
for index, peptide in zip(indices, group['peptides']):
row = []
row.append(group['i'])
row.append(peptide['sequence'])
row.append(peptide['modifications'])
row.append(peptide['protein'])
row.append(' ')
seq = peptide['sequence']
s = ' ' * (left + index) + seq
s = s + ' ' * (max_len - len(s))
for c in s:
if c == ' ':
row.append('.')
else:
row.append(c)
rows.append(row)
rows.append([])
datafile.write_csv(rows, 'cluster.csv')
left = -min(indices)
max_len = 0
for index, seq in zip(indices, sequences):
str_len = left + index + len(seq)
if str_len > max_len:
max_len = str_len
for index, peptide in zip(indices, group['peptides']):
row = []
row.append(group['i'])
row.append(peptide['sequence'])
row.append(peptide['modifications'])
row.append(peptide['protein'])
row.append(' ')
seq = peptide['sequence']
s = ' ' * (left + index) + seq
s = s + ' ' * (max_len - len(s))
for c in s:
if c == ' ':
row.append('.')
else:
row.append(c)
rows.append(row)
rows.append([])
datafile.write_csv(rows, 'cluster.csv')
