def map_pair(seq1, seq1_len, seq2, seq2_len, k):
d1 = 1
d2 = 1
p1 = -1
p2 = -1
c1 = -1
c2 = -1
found1 = 0
for i in range(seq1_len - k + 1):
kit1 = seq1[i:i + k]
kit1_rep = kmer_rep(kit1)
if kit1_rep in kmer_str_dict:
found1 = 1
kit1_info = kmer_str_dict[kit1_rep]
c1 = kit1_info.contig_id
if (kit1_rep == kit1) == kit1_info.sense_in_contig:
p1 = kit1_info.pos_in_contig - i
d1 = 1
else:
p1 = kit1_info.pos_in_contig + k + i
d1 = 0
# print(seq1)
# print(kit1_info.pos_in_contig)
break
else:
pass
if not found1:
return -1
found2 = 0
for i in range(seq2_len - k + 1):
kit2 = seq2[i:i + k]
kit2_rep = kmer_rep(kit2)
if kit2_rep in kmer_str_dict:
found2 = 1
kit2_info = kmer_str_dict[kit2_rep]
c2 = kit2_info.contig_id
if (kit2_rep == kit2) == kit2_info.sense_in_contig:
p2 = kit2_info.pos_in_contig - i
d2 = 1
else:
p2 = kit2_info.pos_in_contig + k + i
d2 = 0
# print(kit2_info.pos_in_contig)
break
else:
pass
if not found2:
return -1
if c1 != c2:
return -1
if d1 and d2:
return -1
if not d1 and not d2:
return -1
if p1 > p2:
return p1 - p2
else:
return p2 - p1
def fix_contigs(k):
perfect_contig_num = 0
for contig_id in range(len(contig_find_trans_list)):
all_ok = 1
# 如前文所言，没有找错的
n_of_kmer_in_curr_contig = contig_list[contig_id].n_of_kmer
for each_trans_info in contig_find_trans_list[contig_id]:
if each_trans_info.start_in_tran != 0 or each_trans_info.stop_in_tran != n_of_kmer_in_curr_contig:
all_ok = 0
# 说明这里有找错的
if all_ok:
perfect_contig_num = perfect_contig_num + 1
else:
tmp_list = []
for each_trans_info in contig_find_trans_list[contig_id]:
tmp_list.append(each_trans_info.start_in_tran)
tmp_list.append(each_trans_info.stop_in_tran)
break_points = sorted(list(set(tmp_list)))
old_contig_seq = contig_list[contig_id].seq
old_trans_info = contig_find_trans_list[contig_id]
new_contig = Contig()
for j in range(len(break_points) - 1):
new_contig.seq = old_contig_seq[
break_points[j]:break_points[j + 1] - break_points[j] + k -
1]
new_contig.n_of_kmer = break_points[j + 1] - break_points[j]
if j == 0:
new_contig.id = contig_id
contig_list[contig_id] = new_contig
else:
new_contig.id = len(contig_list)
contig_list.append(new_contig)
ecs_list.append(-1)
for m in range(len(new_contig.seq) - k + 1):
curr_kmer_in_contig = new_contig.seq[m:m + k]
curr_rep = kmer_rep(curr_kmer_in_contig)
if curr_rep in kmer_str_dict:
kmer_str_dict[curr_rep].contig_id = new_contig.id
kmer_str_dict[curr_rep].pos_in_contig = m
kmer_str_dict[
curr_rep].n_of_kmer_in_contig = new_contig.n_of_kmer
else:
program_stop("ecs.py-4")
new_trans_info_list = []
for each_trans_info in old_trans_info:
if not (each_trans_info.start_in_tran >=
break_points[j + 1] or
each_trans_info.stop_in_tran <= break_points[j]):
new_trans_info = ContigFindTrans()
new_trans_info.sense_in_tran = each_trans_info.sense_in_tran
new_trans_info.tran_id = each_trans_info.tran_id
new_trans_info.start_in_tran = 0
new_trans_info.stop_in_tran = new_contig.n_of_kmer
new_trans_info_list.append(new_trans_info)
if j == 0:
contig_find_trans_list[contig_id] = new_trans_info_list
else:
contig_find_trans_list.append(new_trans_info_list)
def fwstep(kmer):
global fw_step_kmer
fw_count = 0
for fw_kmer in contig_fw(kmer):
curr_rep = kmer_rep(fw_kmer)
if curr_rep in kmer_str_dict:
fw_count = fw_count + 1
curr_fw = fw_kmer
if fw_count > 1:
return 0
if fw_count != 1:
return 0
bw_count = sum(
kmer_rep(tmp) in kmer_str_dict for tmp in contig_bw(curr_fw))
if bw_count > 1:
return 0
if bw_count != 1:
return 0
else:
if curr_fw != kmer:
fw_step_kmer = curr_fw
return 1
else:
return 0
def build_ecs(seqs, k):
contig_find_trans_dict = {}
for i in range(len(contig_list)):
contig_find_trans_dict[i] = []
# 遍历基因序列
for i in range(len(seqs)):
each_ec = [i]
ec_map.append(each_ec)
ec_inv_dict[i] = each_ec
seq = seqs[i]
n_of_kmers_of_seq = len(seq) - k + 1
j = 0
while j < n_of_kmers_of_seq:
curr_kmer = seq[j:j + k]
curr_kmer = str(curr_kmer)
curr_rep = kmer_rep(curr_kmer)
curr_twin = kmer_twin(curr_kmer)
if curr_rep in kmer_str_dict:
curr_contig_id = kmer_str_dict[curr_rep].contig_id
curr_contig = contig_list[curr_contig_id]
kmers_in_contig = []
# 遍历curr_contig里面的每一个kmer
for m in range(len(curr_contig.seq) - k + 1):
curr_kmer_in_contig = curr_contig.seq[m:m + k]
kmers_in_contig.append(curr_kmer_in_contig)
kmers_in_contig_len = len(kmers_in_contig)
contig_find_trans = ContigFindTrans()
contig_find_trans.tran_id = i
if (curr_kmer == curr_rep
) == kmer_str_dict[curr_rep].sense_in_contig:
contig_find_trans.sense_in_tran = 1
contig_find_trans.start_in_tran = kmers_in_contig.index(
curr_kmer)
if kmers_in_contig_len - contig_find_trans.start_in_tran > n_of_kmers_of_seq - j:
# 说明这个contig不来自这个seq
contig_find_trans.stop_in_tran = contig_find_trans.start_in_tran + n_of_kmers_of_seq - j
# [start,stop)这里这个长度肯定是错误的，正好后面要用到
j = n_of_kmers_of_seq
# 到末尾，也就是结束
else:
contig_find_trans.stop_in_tran = kmers_in_contig_len
# contig属于这个seq，而且长度就是contig的长度
j = j + contig_find_trans.stop_in_tran - contig_find_trans.start_in_tran
# 下一个从紧接着当前contig的下一个contig开始
else:
contig_find_trans.sense_in_tran = 0
contig_find_trans.stop_in_tran = kmers_in_contig.index(
curr_twin) + 1
# 因为[start,stop) stop代表的是长度，取不到stop，实际取的是stop-1
if contig_find_trans.stop_in_tran > n_of_kmers_of_seq - j:
contig_find_trans.start_in_tran = contig_find_trans.stop_in_tran - n_of_kmers_of_seq + j
j = n_of_kmers_of_seq
else:
contig_find_trans.start_in_tran = 0
j = j + contig_find_trans.stop_in_tran - contig_find_trans.start_in_tran
contig_find_trans_dict[curr_contig_id].append(
contig_find_trans)
else:
print(curr_rep)
print(j)
program_stop("ecs.py-2")
for i in range(len(contig_list)):
contig_find_trans_list.append(contig_find_trans_dict[i])
fix_contigs(k)
for curr_contig_id in range(len(contig_find_trans_list)):
curr_contig_list = []
for curr_trans_info in contig_find_trans_list[curr_contig_id]:
curr_contig_list.append(curr_trans_info.tran_id)
curr_contig_list = sorted(list(set(curr_contig_list)))
ec = -1
if curr_contig_list in ec_inv_dict.values():
for (key, value) in ec_inv_dict.items():
if value == curr_contig_list:
ec = key
else:
ec = len(ec_inv_dict)
ec_inv_dict[ec] = curr_contig_list
ec_map.append(curr_contig_list)
ecs_list[curr_contig_id] = ec
contig_list[curr_contig_id].ecs_id = ec
for i in range(len(seqs)):
seq = seqs[i]
n_of_kmers_of_seq = len(seq) - k + 1
tmp_str = ""
j = 0
while j < n_of_kmers_of_seq:
curr_kmer = seq[j:j + k]
curr_kmer = str(curr_kmer)
curr_rep = kmer_rep(curr_kmer)
if curr_rep in kmer_str_dict:
curr_kmer_info = kmer_str_dict[curr_rep]
contig_include_trans = ContigIncludeTrans()
contig_include_trans.tran_id = i
contig_include_trans.pos_in_tran = j
contig_include_trans.sense_in_tran = int(((
curr_kmer == curr_rep) == curr_kmer_info.sense_in_contig))
contig_list[curr_kmer_info.contig_id].include_trans.append(
contig_include_trans)
j = j + contig_list[curr_kmer_info.contig_id].n_of_kmer
# if contig_include_trans.sense_in_tran:
# if contig_include_trans.pos_in_tran == 0:
# tmp_str = tmp_str + contig_list[curr_kmer_info.contig_id].seq
# else:
# tmp_str = tmp_str + contig_list[curr_kmer_info.contig_id].seq[k - 1]
# else:
# new_tm_str = str_pair(contig_list[curr_kmer_info.contig_id].seq)
# if contig_include_trans.pos_in_tran == 0:
# tmp_str = tmp_str + new_tm_str
# else:
# tmp_str = tmp_str + new_tm_str[k - 1]
else:
program_stop("ecs.py-3")
def match(curr_read, curr_read_len, k):
# global num_read
# num_read = num_read + 1
# if num_read %2 == 0:
# print(num_read)
v = []
i = 0
again = 0
while i < curr_read_len - k + 1:
kit = curr_read[i:i + k]
kit_rep = kmer_rep(kit)
if kit_rep in kmer_str_dict:
curr_kit_info = kmer_str_dict[kit_rep]
pos = i
v.append((curr_kit_info, pos))
dist = get_dist(int((kit == kit_rep)), curr_kit_info)
if dist > 2:
pos2 = pos + dist
if pos2 + k >= curr_read_len:
# 超界,检查该read的最后一个kmer
pos2 = curr_read_len - k
kit2 = curr_read[pos2:pos2 + k]
kit2_rep = kmer_rep(kit2)
if kit2_rep in kmer_str_dict:
found2 = 0
found2pos = pos + dist
kit2_info = kmer_str_dict[kit2_rep]
if curr_kit_info.contig_id == kit2_info.contig_id:
found2 = 1
found2pos = pos + dist
else:
found2 = 1
found2pos = pos
if found2:
if found2pos >= curr_read_len - k:
v.append((curr_kit_info, curr_read_len - k))
return v
else:
v.append((curr_kit_info, found2pos))
else:
found_middle = 0
if dist > 4:
middle_pos = int((pos + pos2) / 2)
found3pos = pos + dist
kit3 = curr_read[middle_pos:middle_pos + k]
kit3_rep = kmer_rep(kit3)
if kit3_rep in kmer_str_dict:
kit3_info = kmer_str_dict[kit3_rep]
middle_contig_id = kit3_info.contig_id
if middle_contig_id == curr_kit_info.contig_id:
found_middle = 1
found3pos = middle_pos
elif middle_contig_id == kit2_info.contig_id:
found_middle = 1
found3pos = pos + dist
else:
pass
if found_middle:
v.append((kit3_info, found3pos))
if pos2 >= curr_read_len - k:
return v
else:
i = pos2
else:
pass
else:
pass
else:
pass
if not found_middle:
i = i + 1
again = 1
if again:
j = 0
m = 0
while j < curr_read_len and m < curr_read_len:
if j == skip:
j = 0
if j == 0:
again_kit = curr_read[m:m + k]
again_kit_rep = kmer_rep(again_kit)
if again_kit_rep in kmer_str_dict:
again_kit_info = kmer_str_dict[
again_kit_rep]
v.append((again_kit_info, m))
if m == pos2:
again = 0
return v
m = m + 1
j = j + 1
else:
pass
else:
pass
i = i + 1
return v
def build_dbg(seqs, k):
global fw_step_kmer
tmp_kmer_map = set()
# 把序列打断,得到kmer的集合
for seq in seqs:
for i in range(len(seq) - k + 1):
curr_kmer = seq[i:i + k]
curr_kmer = str(curr_kmer)
curr_rep = kmer_rep(curr_kmer)
tmp_kmer_map.add(curr_rep)
# 给每一个kmer带上一个可以存储它信息的类
for each_kmer in tmp_kmer_map:
kmer_info = KmerInfo()
kmer_str_dict[each_kmer] = kmer_info
# 把kmer连起来,找contig
for kmer_key in kmer_str_dict:
curr_kmer = kmer_key
curr_kmer_info = kmer_str_dict[kmer_key]
if curr_kmer_info.contig_id == -1:
self_loop = 0
curr_twin = kmer_twin(curr_kmer)
fw_list = [curr_kmer]
last_kmer = curr_kmer
fw_step_kmer = curr_kmer
while fwstep(fw_step_kmer):
if fw_step_kmer == curr_kmer:
# 该if判断是否成环
self_loop = 1
break
# pass
# example(3,"ACTGAC") (5,"ACCAACCA") (5,"TCTGTCTG") (5,"AACAAACA") (5,"CACACA") (5,"ACACAC")
# print("begin:",begin,"fw:", fw, "t_kmer", self.tmp_kmer, "seqs:", self.seqs)
elif fw_step_kmer == curr_twin:
self_loop = (len(fw_list) > 1)
break
elif fw_step_kmer == kmer_twin(last_kmer):
break
fw_list.append(fw_step_kmer)
last_kmer = fw_step_kmer
fw_step_kmer = curr_twin
bw_list = []
first_kmer = curr_twin
if not self_loop:
while fwstep(fw_step_kmer):
if fw_step_kmer == curr_twin:
break
elif fw_step_kmer == curr_kmer:
break
elif fw_step_kmer == kmer_twin(first_kmer):
break
bw_list.append(fw_step_kmer)
first_kmer = fw_step_kmer
curr_kmer_list = []
for bw in reversed(bw_list):
curr_kmer_list.append(kmer_twin(bw))
for fw in fw_list:
curr_kmer_list.append(fw)
contig_list_len = len(curr_kmer_list)
curr_contig_id = len(contig_list)
contig = Contig()
contig.id = curr_contig_id
contig.n_of_kmer = contig_list_len
# 存储kmer的信息
for j in range(contig_list_len):
each_kmer = curr_kmer_list[j]
tmp_rep = kmer_rep(each_kmer)
if tmp_rep in kmer_str_dict:
kmer_str_dict[tmp_rep].contig_id = curr_contig_id
kmer_str_dict[tmp_rep].pos_in_contig = j
kmer_str_dict[
tmp_rep].n_of_kmer_in_contig = contig_list_len
if tmp_rep == each_kmer:
kmer_str_dict[tmp_rep].sense_in_contig = 1
else:
kmer_str_dict[tmp_rep].sense_in_contig = 0
else:
program_stop("contig.py")
if j == 0:
contig.seq = each_kmer
elif j > 0:
contig.seq = contig.seq + each_kmer[-1]
contig_list.append(contig)
ecs_list.append(-1)