def generate_chromosome(seq,
markov_list,
coord_adjust,
rpt_gen,
mask=False,
max_interval=None,
num_repeats=None,
max_length=None,
limiting_chr=None,
sim_type=0):
"""
Generate a syntehtic sequence with real repeats:
* seq: A sequence (as a string).
* markov_list: List of the k+1 i-th order markov chains (from the markov_gen module).
* coord_adjust: Size of the prefix that has been cut off the template sequence (requiring that
* that .fa.out coordinate be adjusted).
* rpt_gen: A generating function returning the repeat information (created by nextRepeat)
* mask: If true, all repeats will be lower-case. Otherwise, upper case.)
* max_interval: Maximum inter-repeat length.
"""
current_coord = coord_adjust
if max_interval == -1:
max_interval = len(seq)
s = [] # Hold the sequence (in chunks)
fa_out = [] # Hold the new .fa.out file contents (by line)
rpt_count = 0
length = min(len(seq), max_length) if max_length else len(seq)
for chr, start, finish, strand, family, rpt_class, rpt_id, ancestor_seq, modern_seq in rpt_gen:
if limiting_chr and chr not in limiting_chr:
continue
if start >= current_coord:
rpt_count += 1
inter_seq_len = min(start - current_coord, max_interval)
inter_seq = markov_gen.generate_sequence(markov_list,
inter_seq_len)
assert len(inter_seq) == inter_seq_len
s.append(inter_seq)
coord_adjust += max(0, start - current_coord - max_interval)
if sim_type == 0:
rpt_seq = seq[start:finish]
if sim_type == 1:
rpt_seq = re.sub("-", "", ancestor_seq)
elif sim_type == 2:
rpt_seq = "".join([
m if m != '-' else a
for a, m in zip(ancestor_seq, modern_seq) if a != '-'
])
else: # sim_type == 3
rpt_seq = "".join([
a if a != '-' else m
for a, m in zip(ancestor_seq, modern_seq) if m != '-'
])
rpt_seq = "".join([
a if a.upper() in {'A', 'C', 'T', 'G'} else 'N'
for a in rpt_seq
])
s.append(rpt_seq.lower() if mask else rpt_seq.upper())
new_finish = start + len(rpt_seq)
fa_out.append([
chr, start + 1 - coord_adjust, new_finish - coord_adjust,
strand, family, rpt_class, rpt_id
])
coord_adjust += finish - new_finish
if num_repeats and rpt_count == num_repeats:
break
current_coord = finish
if num_repeats:
max_interval = min(1000, max_interval)
tail_length = min(max_interval, length - current_coord)
if tail_length > 0:
s.append(markov_gen.generate_sequence(markov_list, tail_length))
sim_seq = "".join(s)
sim_seq_len = len(sim_seq)
fa_out_str = fa_out_header
for chr, start, finish, strand, family, rpt_class, rpt_id in fa_out:
fa_out_str += fa_out_template.format(chr=chr,
start=start,
finish=finish,
left=sim_seq_len - finish,
strand=strand,
family=family,
rpt_class=rpt_class,
rpt_id=rpt_id)
return sim_seq, fa_out_str
def generate_chromosome(seq, markov_list, chr_start, chr_finish, rpt_gen, mask = False, max_interval = -1, min_interval = 0,num_repeats = None, max_length = None, limiting_chr = None, rep_base_hash = None):
"""
Generate a syntehtic sequence with real repeats:
* seq: A sequence (as a string).
* markov_list: List of the k+1 i-th order markov chains (from the markov_gen module).
* start/finish: Defined the coordinates of our actual template sequence. (We are ignoring anything that occurs before/faster.
* Allows us to cut of a prefix and/or suffix.
* rpt_gen: A generating function returning the repeat information (created by nextRepeat)
* mask: If true, all repeats will be lower-case. Otherwise, upper case.)
* max_interval: Maximum inter-repeat length.
* min_interval: Minimum allowed length of a sequence between repeats. If two repeats are closer than this,
* extend the length.
* max_interval: Minimum allowed length of a sequence between repeats. If two repeats are closer than this,
* cut the length.
"""
last_end = chr_start
if max_interval == -1:
max_interval = len(seq)
sim_seq = "" # Simulated sequence
fa_out = [] # Hold the new .fa.out file contents (by line)
rpt_count = 0 # Count of repeats (so we can quit when we reach num_repeats, if applicable)
for chr, start, finish, strand, family, rpt_class, rpt_id in rpt_gen:
if limiting_chr and chr not in limiting_chr: # Skip if we are on the wrong chromsome
continue
if start >= chr_finish: # Quit if we have gone past the allowed range (repeats are assumed to be sorted by start)
break
if start < chr_start or finish > chr_finish: # Skip if we are outside the allowed range
continue
if start < last_end: # Skip if this repeat overlapped the last one
continue
rpt_count += 1
# Add the next inter-TE sequence
inter_seq_len = max(min_interval, min(start - last_end, max_interval))
sim_seq += markov_gen.generate_sequence(markov_list, inter_seq_len)
# Add the next sequence
if rep_base_hash:
rpt_seq = rep_base_hash[family]
else:
rpt_seq = seq[start:finish]
fa_out.append([chr, len(sim_seq)+1, len(sim_seq) + len(rpt_seq), strand, family, rpt_class, rpt_id]) # Coords adjusted for biologist notation
sim_seq += rpt_seq.lower() if mask else rpt_seq.upper()
if rpt_count == num_repeats:
break
last_end = max(last_end, finish)
# Add final sequence on
final_seq_len = max(min_interval, min(chr_finish - last_end, max_interval))
sim_seq += markov_gen.generate_sequence(markov_list, inter_seq_len)
sim_seq_len = len(sim_seq)
fa_out_str = fa_out_header
for chr, start, finish, strand, family, rpt_class, rpt_id in fa_out:
fa_out_str += fa_out_template.format(chr=chr, start=start, finish=finish, left = sim_seq_len - finish, strand=strand, family=family, rpt_class=rpt_class, rpt_id=rpt_id)
return sim_seq, fa_out_str
def generate_chromosome(seq, markov_list, coord_adjust, rpt_gen, mask = False, max_interval = None, num_repeats = None, max_length = None, limiting_chr = None):
"""
Generate a syntehtic sequence with real repeats:
* seq: A sequence (as a string).
* markov_list: List of the k+1 i-th order markov chains (from the markov_gen module).
* coord_adjust: Size of the prefix that has been cut off the template sequence (requiring that
* that .fa.out coordinate be adjusted).
* rpt_gen: A generating function returning the repeat information (created by nextRepeat)
* mask: If true, all repeats will be lower-case. Otherwise, upper case.)
* max_interval: Maximum inter-repeat length.
"""
current_coord = coord_adjust
if max_interval == -1:
max_interval = len(seq)
s = [] # Hold the sequence (in chunks)
fa_out = [] # Hold the new .fa.out file contents (by line)
rpt_count = 0
length = min(len(seq), max_length) if max_length else len(seq)
debug_sim_len = 0
for chr, start, finish, strand, family, rpt_class, rpt_id in rpt_gen:
if limiting_chr and chr not in limiting_chr:
continue
if start >= current_coord:
rpt_count += 1
inter_seq_len = min(start-current_coord, max_interval)
inter_seq = markov_gen.generate_sequence(markov_list, inter_seq_len)
assert len(inter_seq) == inter_seq_len
s.append(inter_seq)
debug_sim_len += len(inter_seq)
coord_adjust += max(0, start-current_coord-max_interval)
rpt_seq = seq[start:finish]
s.append(rpt_seq.lower() if mask else rpt_seq.upper())
debug_sim_len += len(rpt_seq)
fa_out.append([chr, start+1-coord_adjust, finish-coord_adjust, strand, family, rpt_class, rpt_id])
#fa_out.append(fa_out_template.format(chr=chr, start=start+1-coord_adjust, finish=finish-coord_adjust, strand=strand, family=family, rpt_class=rpt_class, rpt_id=rpt_id))
if num_repeats and rpt_count == num_repeats:
break
current_coord = finish
if num_repeats:
max_interval = min(1000, max_interval)
tail_length = min(max_interval, length-current_coord)
if tail_length > 0:
s.append(markov_gen.generate_sequence(markov_list, tail_length))
sim_seq = "".join(s)
sim_seq_len = len(sim_seq)
fa_out_str = fa_out_header
for chr, start, finish, strand, family, rpt_class, rpt_id in fa_out:
fa_out_str += fa_out_template.format(chr=chr, start=start, finish=finish, left = sim_seq_len - finish, strand=strand, family=family, rpt_class=rpt_class, rpt_id=rpt_id)
return sim_seq, fa_out_str
def generate_chromosome(seq,
markov_list,
chr_start,
chr_finish,
rpt_gen,
mask=False,
max_interval=-1,
min_interval=0,
num_repeats=None,
max_length=None,
limiting_chr=None,
rep_base_hash=None):
"""
Generate a syntehtic sequence with real repeats:
* seq: A sequence (as a string).
* markov_list: List of the k+1 i-th order markov chains (from the markov_gen module).
* start/finish: Defined the coordinates of our actual template sequence. (We are ignoring anything that occurs before/faster.
* Allows us to cut of a prefix and/or suffix.
* rpt_gen: A generating function returning the repeat information (created by nextRepeat)
* mask: If true, all repeats will be lower-case. Otherwise, upper case.)
* max_interval: Maximum inter-repeat length.
* min_interval: Minimum allowed length of a sequence between repeats. If two repeats are closer than this,
* extend the length.
* max_interval: Minimum allowed length of a sequence between repeats. If two repeats are closer than this,
* cut the length.
"""
last_end = chr_start
if max_interval == -1:
max_interval = len(seq)
sim_seq = "" # Simulated sequence
fa_out = [] # Hold the new .fa.out file contents (by line)
rpt_count = 0 # Count of repeats (so we can quit when we reach num_repeats, if applicable)
for chr, start, finish, strand, family, rpt_class, rpt_id in rpt_gen:
if limiting_chr and chr not in limiting_chr: # Skip if we are on the wrong chromsome
continue
if start >= chr_finish: # Quit if we have gone past the allowed range (repeats are assumed to be sorted by start)
break
if start < chr_start or finish > chr_finish: # Skip if we are outside the allowed range
continue
if start < last_end: # Skip if this repeat overlapped the last one
continue
rpt_count += 1
# Add the next inter-TE sequence
inter_seq_len = max(min_interval, min(start - last_end, max_interval))
sim_seq += markov_gen.generate_sequence(markov_list, inter_seq_len)
# Add the next sequence
if rep_base_hash:
rpt_seq = rep_base_hash[family]
else:
rpt_seq = seq[start:finish]
fa_out.append([
chr,
len(sim_seq) + 1,
len(sim_seq) + len(rpt_seq), strand, family, rpt_class, rpt_id
]) # Coords adjusted for biologist notation
sim_seq += rpt_seq.lower() if mask else rpt_seq.upper()
if rpt_count == num_repeats:
break
last_end = max(last_end, finish)
# Add final sequence on
final_seq_len = max(min_interval, min(chr_finish - last_end, max_interval))
sim_seq += markov_gen.generate_sequence(markov_list, inter_seq_len)
sim_seq_len = len(sim_seq)
fa_out_str = fa_out_header
for chr, start, finish, strand, family, rpt_class, rpt_id in fa_out:
fa_out_str += fa_out_template.format(chr=chr,
start=start,
finish=finish,
left=sim_seq_len - finish,
strand=strand,
family=family,
rpt_class=rpt_class,
rpt_id=rpt_id)
return sim_seq, fa_out_str
