def gff3_to_gtf(gff3_file):
dialect = {
'field separator': '; ',
'fmt': 'gtf',
'keyval separator': ' ',
'leading semicolon': False,
'multival separator': ',',
'quoted GFF2 values': True,
'order': ['gene_id', 'transcript_id'],
'repeated keys': False,
'trailing semicolon': True
}
out_file = os.path.splitext(gff3_file)[0] + ".gtf"
if file_exists(out_file):
return out_file
print "Converting %s to %s." % (gff3_file, out_file)
db = gffutils.create_db(gff3_file, ":memory:")
with file_transaction(out_file) as tx_out_file:
with open(tx_out_file, "w") as out_handle:
for feature in DataIterator(db.features_of_type("exon"),
dialect=dialect):
transcript_id = feature["Parent"][0]
gene_id = db[transcript_id]["Parent"][0]
attr = {"transcript_id": transcript_id, "gene_id": gene_id}
attributes = gffutils.attributes.Attributes(attr)
feature.attributes = attributes
print >> out_handle, feature
return out_file
def _output_ncbi_gff3(gff3_file, out_file, dialect):
gene_key = "gene"
id_spec = {"gene": gene_key}
db = gffutils.create_db(gff3_file, ":memory:", id_spec=id_spec)
with file_transaction(out_file) as tx_out_file:
with open(tx_out_file, "w") as out_handle:
for feature in DataIterator(db.features_of_type("exon"),
dialect=dialect):
# Gnomon features are often missing a transcript id
# some malformed features are also missing the gene key
try:
transcript_id = feature["transcript_id"]
except KeyError:
try:
transcript_id = feature[gene_key]
except KeyError:
continue
gene_id = feature[gene_key]
try:
biotype = feature["gene_biotype"]
except KeyError:
biotype = "unknown"
attr = {
"transcript_id": transcript_id,
"gene_id": gene_id,
"gene_biotype": biotype
}
attributes = gffutils.attributes.Attributes(attr)
feature.attributes = attributes
print(feature, file=out_handle, end="")
def _output_gff3(gff3_file, out_file, dialect):
db = gffutils.create_db(gff3_file, ":memory:")
with file_transaction(out_file) as tx_out_file:
with open(tx_out_file, "w") as out_handle:
for feature in DataIterator(db.features_of_type("exon"), dialect=dialect):
transcript_id = feature["Parent"][0]
gene_id = db[transcript_id]["Parent"][0]
attr = {"transcript_id": transcript_id, "gene_id": gene_id}
attributes = gffutils.attributes.Attributes(attr)
feature.attributes = attributes
print(feature, file=out_handle, end="")
def get_trans_records(dbname, trans_seq_dict):
records = DataIterator(dbname)
e = 0
chromosome = []
strand = []
start = []
end = []
trans_id = []
exon_number = []
exon_id = []
gene_id = []
protein_id = []
strand_dict = {}
protein_id_dict = {}
for record in records:
if ('transcript_type' in record.attributes):
if (record.attributes['transcript_id'][0]
in trans_seq_dict.keys()):
strand_dict[record.attributes['transcript_id'][0]] = record[6]
protein_id_dict[record.attributes['transcript_id'][0]]=(record.attributes['protein_id'][0],\
record.attributes['gene_id'][0])
if (record[2] == 'exon'):
chromosome.append(record[0])
strand.append(record[6])
start.append(int(record[3]))
end.append(int(record[4]))
tmp = record.attributes['ID'][0]
flag = tmp.find(':')
tmp = tmp[flag + 1:]
s = tmp.split(':', 1)
if (s[0].find('_') != -1):
flag = s[0].find('_')
s[0] = s[0][:flag]
trans_id.append(s[0])
exon_number.append(int(s[1]))
gene_id.append(record.attributes['gene_id'][0])
exon_id.append(record.attributes['exon_id'][0])
protein_id.append(record.attributes['protein_id'][0])
e += 1
#print(e)
exon_records={'chr':chromosome,'strand':strand,'start':start,'end':end,'trans_id':trans_id,\
'exon_number':exon_number,'exon_id':exon_id,'protein_id':protein_id,'gene_id':gene_id}
trans_records=pd.DataFrame(exon_records,columns=['chr','strand','start','end','trans_id','exon_number',\
'exon_id','protein_id','gene_id'])
trans_records = trans_records.sort_values(by=['trans_id', 'exon_number'],
axis=0)
return trans_records, strand_dict, protein_id_dict
def get_new_sequence(dfname, dbname, rna_db, het, exclude, output_name):
version = get_version(rna_db)
if (version == 'swissprot'):
db = 'sp'
else:
db = version
sequence_dict = {}
rna_seq = SeqIO.parse(rna_db, 'fasta')
##check the correctness of rna-seqs
for correct in rna_seq:
tmp = correct.id
flag = tmp.find("|")
tmp = tmp[flag + 1:]
cds = tmp.find('CDS:')
if (cds == -1):
print("The format of file of parameter -r(--rna) is incorrect!")
return 0
tmp = tmp[cds:]
cds_end = tmp.find('|')
tmp = tmp[tmp.find(':') + 1:cds_end]
split_flag = tmp.find('-')
if (split_flag == -1):
print("The format of file of parameter -r(--rna) is incorrect!")
return 0
rna_seqs = SeqIO.parse(rna_db, 'fasta')
for seq in rna_seqs:
tmp = seq.id
flag = tmp.find("|")
mrna_id = tmp[:flag]
tmp = tmp[flag + 1:]
cds = tmp.find('CDS:')
tmp = tmp[cds:]
cds_end = tmp.find('|')
tmp = tmp[tmp.find(':') + 1:cds_end]
split_flag = tmp.find('-')
coding_start = int(tmp[:split_flag])
coding_end = int(tmp[split_flag + 1:])
sequence_dict[mrna_id] = (coding_start, seq.seq, coding_end)
print("sequence_dict ready")
records = DataIterator(dbname)
strand_dict = {}
protein_id_dict = {}
for record in records:
if (record[2] == 'transcript'):
if ('transcript_type' in record.attributes):
if (record.attributes['transcript_id'][0]
in sequence_dict.keys()):
strand_dict[record.attributes['transcript_id']
[0]] = record[6]
protein_id_dict[record.attributes['transcript_id'][0]]=(record.attributes['protein_id'][0],\
record.attributes['gene_id'][0])
print("protein_id_dict ready")
df = pd.read_csv(dfname, sep='\t', header=None)
change_df = extract_transcript_change(df)
trans_index_dict = {}
for i in range(0, change_df.shape[0]):
if (change_df.iloc[i]['mrna'] not in trans_index_dict.keys()):
trans_index_dict[change_df.iloc[i]['mrna']] = [i]
else:
trans_index_dict[change_df.iloc[i]['mrna']].append(i)
my_seqs = []
k_cnt = 0
hom_only_cnt = 0
hom_het_cnt = 0
het_only_cnt = 0
original_cnt = 0
random_cnt = 0
for k in trans_index_dict.keys():
if (k in protein_id_dict.keys()):
k_cnt += 1
if (k_cnt % 1000 == 0):
print(k_cnt)
pid = protein_id_dict[k][0]
gid = protein_id_dict[k][1]
if (strand_dict[k] == '+'):
transcript = str(sequence_dict[k][1])
else:
tmp = Seq(str(sequence_dict[k][1]),
IUPAC.ambiguous_dna).complement()
transcript = str(tmp)
coding_start = int(sequence_dict[k][0]) - 1
coding_end = int(sequence_dict[k][2]) - 1
transcript = transcript[coding_start:coding_end + 1]
shift = 0
des = ""
het_list = []
hom_position_list = []
for i in trans_index_dict[k]:
if (change_df.iloc[i]['snp_type'] == 'hom'):
hom_position_list.append(
(int(change_df.iloc[i]['c_start']),
change_df.iloc[i]['mutation_type']))
# if(change_df.iloc[i]['mutation_type']=='snv'):
# transcript=change_seq(transcript,int(change_df.iloc[i]['c_start'])+shift,\
# int(change_df.iloc[i]['c_end'])+shift,change_df.iloc[i]['c_content'],'snv')
# des+="snv:"+str(change_df.iloc[i]['c_start'])+change_df.iloc[i]['c_content']+'_'
# else:
if (strand_dict[k] == '-'):
transcript=change_seq(transcript,int(change_df.iloc[i]['c_start'])+shift,\
int(change_df.iloc[i]['c_end'])+shift,str(Seq(str(change_df.iloc[i]['c_content'])).complement()),\
change_df.iloc[i]['mutation_type'])
else:
transcript=change_seq(transcript,int(change_df.iloc[i]['c_start'])+shift,\
int(change_df.iloc[i]['c_end'])+shift,change_df.iloc[i]['c_content'],\
change_df.iloc[i]['mutation_type'])
des+=change_df.iloc[i]['mutation_type']+":"+str(change_df.iloc[i]['c_start'])+'-'+\
str(change_df.iloc[i]['c_end'])+str(change_df.iloc[i]['c_content'])+'_'
if (change_df.iloc[i]['mutation_type'].find('del') != -1):
shift -= (int(change_df.iloc[i]['c_end']) -
int(change_df.iloc[i]['c_start']) + 1)
elif (change_df.iloc[i]['mutation_type'].find('ins') !=
-1):
shift += (int(change_df.iloc[i]['c_end']) -
int(change_df.iloc[i]['c_start']) + 1)
else:
het_list.append(i)
if (len(hom_position_list) != 0):
new_sequence = ""
new_des = ""
for p in hom_position_list:
flag = des.find(str(p[0]))
tmp = des[flag:]
flag = tmp.find('_')
new_des += str(p[1]) + ":" + tmp[:flag + 1]
new_sequence = transcript
if (len(new_sequence) != 0):
new_sequence = new_sequence[shift:]
if (strand_dict[k] == '+'):
new_seq = str(
Seq(str(new_sequence),
IUPAC.ambiguous_dna).transcribe().translate(
to_stop=True))
else:
new_seq = str(
Seq(str(new_sequence), IUPAC.ambiguous_dna).
complement().transcribe().translate(to_stop=True))
while (new_seq.find('None') != -1):
new_seq = new_seq.replace('None', '')
my_seqs.append(SeqRecord(Seq(str(new_seq),IUPAC.protein),\
id=db+'|'+pid+'|'+k+'|'+gid+'_0:'+str(int(sequence_dict[k][0])+shift)+'-'+str(int(sequence_dict[k][2])+shift)+'_'+new_des,\
description=new_des))
hom_only_cnt += 1
else:
if (exclude == False):
new_sequence = transcript
if (strand_dict[k] == '+'):
new_seq = str(
Seq(str(new_sequence),
IUPAC.ambiguous_dna).transcribe().translate(
to_stop=True))
else:
new_seq = str(
Seq(str(new_sequence), IUPAC.ambiguous_dna).
complement().transcribe().translate(to_stop=True))
my_seqs.append(SeqRecord(Seq(str(new_seq),IUPAC.protein),\
id=db+'|'+pid+'|'+k+'|'+gid+'_0:'+str(int(sequence_dict[k][0]))+'-'+str(int(sequence_dict[k][2]))+'_no_variant',\
description="no variant"))
original_cnt += 1
if (het == 1):
coding_start = int(sequence_dict[k][0]) + shift - 1
coding_end = int(sequence_dict[k][2]) + shift - 1
count = int(len(transcript) / 900)
if (len(transcript) <= 900):
count = 1
cnt = 0
for l in range(0, count):
l = l * 900
start = l
if (start + 1799 < len(transcript)):
stop = start + 1799
else:
stop = len(transcript) - 1
het_number = len(het_list)
for n in range(0, het_number):
new_sequence = ""
new_des = ""
if ((int(change_df.iloc[het_list[n]]['c_start']) >=
start) &
(int(change_df.iloc[het_list[n]]['c_start']) <=
stop)):
if (strand_dict[k] == '-'):
new_sequence=change_seq(transcript,int(change_df.iloc[het_list[n]]['c_start'])+shift,\
int(change_df.iloc[het_list[n]]['c_end'])+shift,\
str(Seq(str(change_df.iloc[het_list[n]]['c_content'])).complement()),\
change_df.iloc[het_list[n]]['mutation_type'])
else:
new_sequence=change_seq(transcript,int(change_df.iloc[het_list[n]]['c_start'])+shift,\
int(change_df.iloc[het_list[n]]['c_end'])+shift,change_df.iloc[het_list[n]]['c_content'],\
change_df.iloc[het_list[n]]['mutation_type'])
new_des=des+change_df.iloc[het_list[n]]['mutation_type']+":"+\
str(change_df.iloc[het_list[n]]['c_start'])+'-'+str(change_df.iloc[het_list[n]]['c_end'])+\
str(change_df.iloc[het_list[n]]['c_content'])
if (len(new_sequence) != 0):
new_sequence = new_sequence[start:stop + 1]
if (strand_dict[k] == '+'):
new_seq = str(
Seq(str(new_sequence), IUPAC.ambiguous_dna
).transcribe().translate(to_stop=True))
else:
new_seq = str(
Seq(str(new_sequence),
IUPAC.ambiguous_dna).complement().
transcribe().translate(to_stop=True))
cnt += 1
while (new_seq.find('None') != -1):
new_seq = new_seq.replace('None', '')
my_seqs.append(SeqRecord(Seq(str(new_seq),IUPAC.protein),\
id=db+'|'+pid+'|'+k+'|'+gid+'_'+str(cnt)+':'+str(start+1)+'-'+str(stop+1)+'_'+new_des,\
description=new_des))
if (len(hom_position_list) != 0):
hom_het_cnt += 1
else:
het_only_cnt += 1
if (change_df.iloc[het_list[n]]['mutation_type'] ==
'snv'):
random_seq, random_des = generate_random_SNV_site(
strand_dict[k], k,
transcript[start:stop + 1],
int(change_df.iloc[het_list[n]]['c_start'])
- start)
my_seqs.append(SeqRecord(Seq(str(random_seq),IUPAC.protein),\
id=db+'|'+pid+'|'+k+'|'+gid+'_'+str(cnt)+':'+str(start+1)+'-'+str(stop+1)+'_'+des+random_des,\
description=des+random_des))
random_cnt += 1
else:
random_seq,random_des=generate_random_fs(strand_dict[k],transcript[start:stop+1],\
change_df.iloc[het_list[n]]['mutation_type'],\
len(change_df.iloc[het_list[n]]['c_content']),\
int(change_df.iloc[het_list[n]]['c_start'])-start)
# if(len(random_seq)!=0):
my_seqs.append(SeqRecord(Seq(str(random_seq),IUPAC.protein),\
id=db+'|'+pid+'|'+k+'|'+gid+'_'+str(cnt)+':'+str(start+1)+'-'+str(stop+1)+'_'+des+random_des,\
description=des+random_des))
random_cnt += 1
if (exclude == False):
for key in sequence_dict.keys():
if (key not in trans_index_dict.keys()):
pid = protein_id_dict[key][0]
gid = protein_id_dict[key][1]
coding_start = int(sequence_dict[key][0]) - 1
coding_end = int(sequence_dict[key][2]) - 1
new_sequence = sequence_dict[key][1][coding_start:coding_end +
1]
new_seq = str(
Seq(str(new_sequence),
IUPAC.ambiguous_dna).transcribe().translate(
to_stop=True))
my_seqs.append(SeqRecord(Seq(str(new_seq),IUPAC.protein),\
id=db+'|'+pid+'|'+key+'|'+gid+'_0:'+str(coding_start+1)+'-'+str(coding_end+1)+'_no_variant',\
description="no variant"))
original_cnt += 1
print("The number of proteins related is " + str(k_cnt))
print("The number of sequences generated is " + str(len(my_seqs)))
handle = open(output_name + ".fasta", "w")
for sequence in my_seqs:
SeqIO.write(sequence, handle, "fasta")
print("The number of sequences containing hom only is " +
str(hom_only_cnt))
print("The number of sequences containing het only is " +
str(het_only_cnt))
print("The number of mixed sequences is " + str(hom_het_cnt))
print("The number of original sequences is " + str(original_cnt))
print("The number of random sequences is " + str(random_cnt))
type=str,
help=
'A gff file containing intervals within which gc content can be determined.'
)
conf = ap.parse_args()
#-----------------------------------------------------
# Step 2
# Identify the gc content of features in the gff file
#-----------------------------------------------------
genome_file = conf.genome
gff_file = conf.gff
for feature in DataIterator(gff_file):
contig_id = str(feature.seqid)
feat_start = str(feature.start)
feat_stop = str(feature.stop)
sequence = feature.sequence(genome_file)
g_count = sequence.count('G')
c_count = sequence.count('C')
n_count = sequence.count('N')
gc_count = float(g_count + c_count)
seq_len = int(len(sequence) - n_count)
gc_frac = np.divide(gc_count, seq_len)
gc_perc = int(np.round_(np.multiply(gc_frac, 100), decimals=0, out=None))
outline = [contig_id, feat_start, feat_stop, str(gc_perc)]
print("\t".join(outline))
def get_new_sequence(dfname, dbname, rna_db, protein_db, dataset_name):
version = get_version(protein_db)
if (version == 'swissprot'):
db = 'sp'
else:
db = version
protein_coding_list = get_protein_coding_list_from_db(protein_db)
records = DataIterator(dbname)
strand_dict = {}
protein_id_dict = {}
for record in records:
if (record[2] == 'transcript'):
if ('transcript_type' in record.attributes):
if (record.attributes['transcript_type'][0] == 'protein_coding'
):
if (record.attributes['transcript_id'][0]
in protein_coding_list):
strand_dict[record.attributes['transcript_id']
[0]] = record[6]
protein_id_dict[record.attributes['transcript_id'][0]]=(record.attributes['protein_id'][0],\
record.attributes['gene_id'][0])
print("protein_id_dict ready")
sequence_dict = {}
rna_seqs = SeqIO.parse(rna_db, 'fasta')
##check the correctness of rna-seqs
for correct in rna_seqs:
tmp = correct.id
flag = tmp.find("|")
mrna_id = tmp[:flag]
tmp = tmp[flag + 1:]
cds = tmp.find('CDS:')
if (cds == -1):
print("The format of file of parameter -r(--rna) is incorrect!")
return 0
tmp = tmp[cds:]
cds_end = tmp.find('|')
tmp = tmp[tmp.find(':') + 1:cds_end]
split_flag = tmp.find('-')
if (split_flag == -1):
print("The format of file of parameter -r(--rna) is incorrect!")
return 0
coding_start = int(tmp[:split_flag])
coding_end = int(tmp[split_flag + 1:])
for seq in rna_seqs:
tmp = seq.id
flag = tmp.find("|")
mrna_id = tmp[:flag]
tmp = tmp[flag + 1:]
cds = tmp.find('CDS:')
tmp = tmp[cds:]
cds_end = tmp.find('|')
tmp = tmp[tmp.find(':') + 1:cds_end]
split_flag = tmp.find('-')
coding_start = int(tmp[:split_flag])
coding_end = int(tmp[split_flag + 1:])
sequence_dict[mrna_id] = (coding_start, seq.seq, coding_end)
print("sequence_dict ready")
df = pd.read_csv(dfname, sep='\t', header=None)
change_df = extract_transcript_change(df)
trans_index_dict = {}
for i in range(0, change_df.shape[0]):
if (change_df.iloc[i]['mrna'] not in trans_index_dict.keys()):
trans_index_dict[change_df.iloc[i]['mrna']] = [i]
else:
trans_index_dict[change_df.iloc[i]['mrna']].append(i)
my_seqs = []
k_cnt = 0
hom_cnt = 0
het_cnt = 0
for k in trans_index_dict.keys():
if (k in protein_id_dict.keys()):
k_cnt += 1
# if(k_cnt%1000==0):
# print(k_cnt)
pid = protein_id_dict[k][0]
gid = protein_id_dict[k][1]
if (strand_dict[k] == '+'):
transcript = str(sequence_dict[k][1])
else:
tmp = Seq(str(sequence_dict[k][1]),
IUPAC.ambiguous_dna).complement()
transcript = str(tmp)
coding_start = int(sequence_dict[k][0]) - 1
coding_end = int(sequence_dict[k][2]) - 1
transcript = transcript[coding_start:coding_end + 1]
shift = 0
des = ""
het_list = []
hom_position_list = []
for i in trans_index_dict[k]:
if (change_df.iloc[i]['snp_type'] == 'hom'):
hom_cnt += 1
hom_position_list.append(
(int(change_df.iloc[i]['c_start']),
change_df.iloc[i]['mutation_type']))
# if(change_df.iloc[i]['mutation_type']=='snv'):
# transcript=change_seq(transcript,int(change_df.iloc[i]['c_start'])+shift,\
# int(change_df.iloc[i]['c_end'])+shift,change_df.iloc[i]['c_content'],'snv')
# des+="snv:"+str(change_df.iloc[i]['c_start'])+change_df.iloc[i]['c_content']+'_'
# else:
if (strand_dict[k] == '-'):
transcript=change_seq(transcript,int(change_df.iloc[i]['c_start'])+shift,\
int(change_df.iloc[i]['c_end'])+shift,str(Seq(str(change_df.iloc[i]['c_content'])).complement()),\
change_df.iloc[i]['mutation_type'])
else:
transcript=change_seq(transcript,int(change_df.iloc[i]['c_start'])+shift,\
int(change_df.iloc[i]['c_end'])+shift,change_df.iloc[i]['c_content'],\
change_df.iloc[i]['mutation_type'])
des+=change_df.iloc[i]['mutation_type']+":"+str(change_df.iloc[i]['c_start'])+'-'+\
str(change_df.iloc[i]['c_end'])+str(change_df.iloc[i]['c_content'])+'_'
if (change_df.iloc[i]['mutation_type'].find('del') != -1):
shift -= (int(change_df.iloc[i]['c_end']) -
int(change_df.iloc[i]['c_start']) + 1)
elif (change_df.iloc[i]['mutation_type'].find('ins') !=
-1):
shift += (int(change_df.iloc[i]['c_end']) -
int(change_df.iloc[i]['c_start']) + 1)
else:
het_list.append(i)
het_cnt += 1
coding_start = int(sequence_dict[k][0]) + shift - 1
coding_end = int(sequence_dict[k][2]) + shift - 1
count = int(len(transcript) / 900)
if (len(transcript) <= 900):
count = 1
cnt = 0
for l in range(0, count):
l = l * 900
start = l
if (start + 1799 < len(transcript)):
stop = start + 1799
else:
stop = len(transcript) - 1
# while(coding_start>stop):
# l=l+1
# start=l*1800
# if(l+1799<len(transcript)):
# stop=l+1799
# else:
# stop=len(transcript)-1
# #if(start>coding_end+shift)
if (len(hom_position_list) != 0):
new_sequence = ""
new_des = ""
for p in hom_position_list:
if ((p[0] - 1 >= start) & (p[0] - 1 <= stop)):
flag = des.find(str(p[0]))
tmp = des[flag:]
flag = tmp.find('_')
new_des += str(p[1]) + ":" + tmp[:flag + 1]
new_sequence = transcript
if (len(new_sequence) != 0):
new_sequence = new_sequence[start:stop + 1]
if (strand_dict[k] == '+'):
new_seq = str(
Seq(str(new_sequence), IUPAC.ambiguous_dna).
transcribe().translate(to_stop=True))
else:
new_seq = str(
Seq(str(new_sequence),
IUPAC.ambiguous_dna).complement().
transcribe().translate(to_stop=True))
cnt += 1
while (new_seq.find('None') != -1):
new_seq = new_seq.replace('None', '')
my_seqs.append(SeqRecord(Seq(str(new_seq),IUPAC.protein),\
id=db+'|'+pid+'|'+k+'|'+gid+'_'+str(cnt)+':'+str(start+1)+'-'+str(stop+1),\
description=new_des))
het_number = len(het_list)
for n in range(0, het_number):
new_sequence = ""
new_des = ""
if ((int(change_df.iloc[het_list[n]]['c_start']) >= start)
&
(int(change_df.iloc[het_list[n]]['c_start']) <= stop)):
if (strand_dict[k] == '-'):
new_sequence=change_seq(transcript,int(change_df.iloc[het_list[n]]['c_start'])+shift,\
int(change_df.iloc[het_list[n]]['c_end'])+shift,\
str(Seq(str(change_df.iloc[het_list[n]]['c_content'])).complement()),\
change_df.iloc[het_list[n]]['mutation_type'])
else:
new_sequence=change_seq(transcript,int(change_df.iloc[het_list[n]]['c_start'])+shift,\
int(change_df.iloc[het_list[n]]['c_end'])+shift,change_df.iloc[het_list[n]]['c_content'],\
change_df.iloc[het_list[n]]['mutation_type'])
new_des=des+change_df.iloc[het_list[n]]['mutation_type']+":"+\
str(change_df.iloc[het_list[n]]['c_start'])+'-'+str(change_df.iloc[het_list[n]]['c_end'])+\
str(change_df.iloc[het_list[n]]['c_content'])
if (len(new_sequence) != 0):
new_sequence = new_sequence[start:stop + 1]
if (strand_dict[k] == '+'):
new_seq = str(
Seq(str(new_sequence), IUPAC.ambiguous_dna).
transcribe().translate(to_stop=True))
else:
new_seq = str(
Seq(str(new_sequence),
IUPAC.ambiguous_dna).complement().
transcribe().translate(to_stop=True))
cnt += 1
while (new_seq.find('None') != -1):
new_seq = new_seq.replace('None', '')
my_seqs.append(SeqRecord(Seq(str(new_seq),IUPAC.protein),\
id=db+'|'+pid+'|'+k+'|'+gid+'_'+str(cnt)+':'+str(start+1)+'-'+str(stop+1),\
description=new_des))
print("The number of proteins related is " + str(k_cnt))
print("The number of sequences generated is " + str(len(my_seqs)))
print("The number of homozygous is " + str(hom_cnt))
print("The number of heterozygous is " + str(het_cnt))
#return my_seqs
handle = open(dataset_name + "_all_mutation_" + version + ".fasta", "w")
hom_seq = 0
het_seq = 0
hom_het_seq = 0
for sequence in my_seqs:
if (str(sequence.description)[-1] == '_'):
hom_seq += 1
else:
if (str(sequence.description).find('_') == -1):
het_seq += 1
else:
hom_het_seq += 1
SeqIO.write(sequence, handle, "fasta")
print("The number of homozygous sequences is " + str(hom_seq))
print("The number of heterozygous sequences is " + str(het_seq))
print("The number of mixed sequences is " + str(hom_het_seq))
def read_taxon_id(run_folder):
"""
Search for Taxon ID in genbank or GFF files.
For GenBank file searc for ''taxon:' key in 'db_xref' qualifier.
For GFF file search for 'taxon' in dbxref feature.
Args:
run_folder (str): path to the input folder
"""
taxon_ids = {}
for input_folder in os.listdir(run_folder):
input_folder_path = os.path.join(run_folder, input_folder)
for input_file in os.listdir(input_folder_path):
if '.gbk' in input_file:
gbk_pathname = os.path.join(input_folder_path, input_file)
# Take the species name and the taxon id from the genbank file.
with open(gbk_pathname, "r") as gbk:
# Take the first record of the genbank (first contig/chromosome) to retrieve the species name.
first_seq_record = next(SeqIO.parse(gbk, "genbank"))
# Take the source feature of the first record.
# This feature contains the taxon ID in the db_xref qualifier.
src_features = [
feature for feature in first_seq_record.features
if feature.type == "source"
]
for src_feature in src_features:
try:
src_dbxref_qualifiers = src_feature.qualifiers[
'db_xref']
for src_dbxref_qualifier in src_dbxref_qualifiers:
if 'taxon:' in src_dbxref_qualifier:
taxon_id = src_dbxref_qualifier.replace(
'taxon:', '')
except KeyError:
logger.info(
'No taxon ID in the Genbank {0} In the FEATURES source you must have: /db_xref="taxon:taxonid" Where taxonid is the Id of your organism. You can find it on the NCBI.'
.format(gbk_pathname))
elif '.gff' in input_file:
gff_pathname = os.path.join(input_folder_path, input_file)
# Instead of parsing and creating a database from the GFF, parse the file and extract the first region feature.
try:
region_feature = [
feature for feature in DataIterator(gff_pathname)
if feature.featuretype == 'region'
][0]
except IndexError:
raise IndexError(
'No region feature in the GFF file of {0}, GFF file must have region features.'
.format(input_folder))
try:
region_feature.attributes['Dbxref']
except KeyError:
raise KeyError(
'No Dbxref in GFF file of {0} GFF file must have a ;Dbxref=taxon:taxonid; in the region feature.'
.format(input_folder))
for dbxref in region_feature.attributes['Dbxref']:
if 'taxon' in dbxref:
taxon_id = dbxref.split('taxon:')[1]
elif '.pf' in input_file:
logger.info(
'No taxon ID associated to a PathoLogic Format. {0} will have a missing taxon_id'
.format(input_folder))
taxon_id = "missing"
taxon_ids[input_folder] = taxon_id
return taxon_ids
def create_flats_and_lisp(run_folder, taxon_file):
"""
Read Genbank/GFF/PF files and create Pathway Tools needed file.
Create also a lisp file to create flat files from Pathway tools results.
The name of the PGDB created by Pathway Tools will be the name of the species with '_' instead of space.
Create organism-params.dat:
ID pgdb_id
STORAGE FILE
NCBI-TAXON-ID taxon_id
NAME species_name
Create genetic-elements.dats:
NAME
ANNOT-FILE gbk_name
//
Create flat_files_creation.lisp:
(in-package :ecocyc)
(select-organism :org-id 'pgdb_id)
(create-flat-files-for-current-kb)
Args:
run_folder (str): ID of a species of the input folder
taxon_file (bool): Boolean indicating if a taxon_file must be used
Returns:
list: boolean list, True if all files have been created
"""
# Look for a Genbank/GFF files in the run folder.
# PGDB ID corresponds to the name of the species folder.
pgdb_id = os.path.basename(run_folder)
gbk_name = pgdb_id + ".gbk"
gbk_pathname = os.path.join(run_folder, gbk_name)
gbff_name = pgdb_id + ".gbff"
gbff_pathname = os.path.join(run_folder, gbff_name)
gff_name = pgdb_id + ".gff"
gff_pathname = os.path.join(run_folder, gff_name)
organism_dat = os.path.join(run_folder, 'organism-params.dat')
genetic_dat = os.path.join(run_folder, 'genetic-elements.dat')
lisp_pathname = os.path.join(run_folder, 'flat_files_creation.lisp')
fasta_extensions = ['.fasta', '.fsa']
taxon_id = ""
taxon_error = False
species_name = ""
taxon_datas = {}
if os.path.isfile(gbk_pathname) or os.path.isfile(gbff_pathname):
if os.path.isfile(gbk_pathname):
input_name = gbk_name
input_path = gbk_pathname
else:
input_name = gbff_name
input_path = gbff_pathname
# Take the species name and the taxon id from the genbank file.
with open(input_path, "r") as gbk:
# Take the first record of the genbank (first contig/chromosome) to retrieve the species name.
try:
first_seq_record = next(SeqIO.parse(gbk, "genbank"))
except StopIteration:
logger.critical(
'Issue with the genbank {0}, it can be empty or malformatted.'
.format(input_path))
return None
try:
species_name = first_seq_record.annotations['organism']
except KeyError:
logger.critical(
'No organism in the Genbank {0} In the SOURCE you must have: ORGANISM Species name'
.format(pgdb_id))
return None
# Take the source feature of the first record.
# This feature contains the taxon ID in the db_xref qualifier.
src_features = [
feature for feature in first_seq_record.features
if feature.type == "source"
]
for src_feature in src_features:
if 'db_xref' in src_feature.qualifiers:
src_dbxref_qualifiers = src_feature.qualifiers['db_xref']
for src_dbxref_qualifier in src_dbxref_qualifiers:
if 'taxon:' in src_dbxref_qualifier:
taxon_id = src_dbxref_qualifier.replace(
'taxon:', '')
if not taxon_id:
logger.info(
'No taxon ID in the Genbank {0} In the FEATURES source you must have: /db_xref="taxon:taxonid" Where taxonid is the Id of your organism. You can find it on the NCBI.'
.format(gbk_pathname))
logger.info('Try to look in the taxon_id.tsv file')
taxon_error, taxon_id, taxon_datas = extract_taxon_id(
run_folder, pgdb_id, taxon_id, taxon_file)
if taxon_file:
taxon_error, taxon_id, taxon_datas = extract_taxon_id(
run_folder, pgdb_id, taxon_id, taxon_file)
elif os.path.isfile(gff_pathname):
input_name = gff_name
# Check if there is a fasta file.
gff_fasta = None
for fasta_extension in fasta_extensions:
fasta_input_name = input_name.replace('.gff', fasta_extension)
fasta_path = os.path.join(run_folder, fasta_input_name)
if os.path.exists(fasta_path):
gff_fasta = fasta_input_name
if not gff_fasta:
logger.critical(
'No fasta file (.fasta or .fsa) with the GFF of {0}'.format(
pgdb_id))
return None
# Instead of parsing and creating a database from the GFF, parse the file and extract the first region feature.
try:
region_feature = [
feature for feature in DataIterator(gff_pathname)
if feature.featuretype == 'region'
][0]
except IndexError:
logger.critical(
'No region feature in the GFF file of {0}, GFF file must have region features.'
.format(pgdb_id))
return None
try:
region_feature.attributes['Dbxref']
except KeyError:
logger.critical(
'No Dbxref in GFF file of {0} GFF file must have a ;Dbxref=taxon:taxonid; in the region feature.'
.format(pgdb_id))
for dbxref in region_feature.attributes['Dbxref']:
if 'taxon' in dbxref:
taxon_id = dbxref.split('taxon:')[1]
if not taxon_id or taxon_file:
if not taxon_id:
logger.info(
'Missing "taxon:" in GFF file of {0} GFF file must have a ;Dbxref=taxon:taxonid; in the region feature.'
.format(pgdb_id))
logger.info('Try to look in the taxon_id.tsv file')
taxon_error, taxon_id, taxon_datas = extract_taxon_id(
run_folder, pgdb_id, taxon_id, taxon_file)
# Look for PF files.
elif all([
True for species_file in os.listdir(run_folder)
if '.pf' in species_file or '.fasta' in species_file
or '.fsa' in species_file
]):
for species_file in os.listdir(run_folder):
if '.pf' in species_file:
# Check if there is a fasta file.
pf_fasta = None
for fasta_extension in fasta_extensions:
fasta_species_name = species_file.replace(
'.pf', fasta_extension)
fasta_path = os.path.join(run_folder, fasta_species_name)
if os.path.exists(fasta_path):
pf_fasta = fasta_species_name
if not pf_fasta:
logger.critical(
'No fasta file (.fasta or .fsa) with the Pathologic file of {0}, this could lead to warnings in Pathway Tools.'
.format(pgdb_id))
taxon_error, taxon_id, taxon_datas = extract_taxon_id(
run_folder, pgdb_id, taxon_id, taxon_file)
if taxon_error == True:
logger.critical('Issue with taxon ID of {0}.'.format(run_folder))
return None
# Create the organism-params dat file.
with open(organism_dat, 'w', encoding='utf-8') as organism_file:
organism_writer = csv.writer(organism_file,
delimiter='\t',
lineterminator='\n')
organism_writer.writerow(['ID', pgdb_id])
organism_writer.writerow(['STORAGE', "FILE"])
organism_writer.writerow(['NCBI-TAXON-ID', taxon_id])
organism_writer.writerow(['NAME', species_name])
if 'reference_pgdbs' in taxon_datas:
for reference_pgdb in taxon_datas['reference_pgdbs']:
organism_writer.writerow(['REF-ORGID', reference_pgdb])
# Create the genetic-elements dat file.
with open(genetic_dat, 'w', encoding='utf-8') as genetic_file:
if os.path.isfile(gff_pathname) or os.path.isfile(
gbk_pathname) or os.path.isfile(gbff_pathname):
genetic_writer = csv.writer(genetic_file,
delimiter='\t',
lineterminator='\n')
genetic_writer.writerow(['NAME', ''])
genetic_writer.writerow(['ANNOT-FILE', input_name])
if os.path.isfile(gff_pathname):
genetic_writer.writerow(['SEQ-FILE', gff_fasta])
if 'circular' in taxon_datas:
circular = taxon_datas['circular']
genetic_writer.writerow(['CIRCULAR?', circular])
if 'element_type' in taxon_datas:
element_type = taxon_datas['element_type']
genetic_writer.writerow(['TYPE', element_type])
if 'codon_table' in taxon_datas:
codon_table = taxon_datas['codon_table']
genetic_writer.writerow(['CODON-TABLE', codon_table])
genetic_writer.writerow(['//'])
elif all([
True for species_file in os.listdir(run_folder)
if '.pf' in species_file or '.fasta' in species_file
or '.fsa' in species_file
]):
genetic_writer = csv.writer(genetic_file,
delimiter='\t',
lineterminator='\n')
for species_file in os.listdir(run_folder):
if '.pf' in species_file:
species_file_name = os.path.splitext(species_file)[0]
genetic_writer.writerow(
['NAME', species_file.replace('.pf', '')])
genetic_writer.writerow(
['ID', species_file.replace('.pf', '')])
genetic_writer.writerow(['ANNOT-FILE', species_file])
fasta_path = os.path.join(
run_folder, species_file.replace('.pf', '.fasta'))
fsa_path = os.path.join(
run_folder, species_file.replace('.pf', '.fsa'))
if os.path.exists(fasta_path):
genetic_writer.writerow([
'SEQ-FILE',
species_file.replace('.pf', '.fasta')
])
elif os.path.exists(fsa_path):
genetic_writer.writerow(
['SEQ-FILE',
species_file.replace('.pf', '.fsa')])
if species_file_name in taxon_datas:
if 'circular' in taxon_datas[species_file_name]:
circular = taxon_datas[species_file_name][
'circular']
genetic_writer.writerow(['CIRCULAR?', circular])
if 'element_type' in taxon_datas[species_file_name]:
element_type = taxon_datas[species_file_name][
'element_type']
genetic_writer.writerow(['TYPE', element_type])
if 'codon_table' in taxon_datas[species_file_name]:
codon_table = taxon_datas[species_file_name][
'codon_table']
genetic_writer.writerow(
['CODON-TABLE', codon_table])
else:
if 'circular' in taxon_datas:
circular = taxon_datas['circular']
genetic_writer.writerow(['CIRCULAR?', circular])
if 'element_type' in taxon_datas:
element_type = taxon_datas['element_type']
genetic_writer.writerow(['TYPE', element_type])
if 'codon_table' in taxon_datas:
codon_table = taxon_datas['codon_table']
genetic_writer.writerow(
['CODON-TABLE', codon_table])
genetic_writer.writerow(['//'])
# Create the lisp script.
check_lisp_file = create_flat_creation_script(pgdb_id, lisp_pathname)
return all([
os.path.isfile(organism_dat),
os.path.isfile(genetic_dat), check_lisp_file
])