Myujth

I have text file called test.txt. From test.txt, I want to grab the lines that start with >lcl then to extract values after locus tag and within close bracket]. I want to do the same thing for values after location. The result I want is shown below. How can I do this in python?

desired result

SS1G_08319 <504653..>506706
SS1G_12233 complement(<502136..>503461)
SS1G_02099 <2692251..>2693298
SS1G_05227 complement(<1032740..>1033620)

test.txt

>lcl|NW_001820825.1_gene_208 [locus_tag=SS1G_08319] [db_xref=GeneID:5486863] [partial=5',3'] [location=<504653..>506706] [gbkey=Gene]
ATGGGCAAAGCTTCTAGGAATAAGACGAAGCATCGCGCTGATCCTACCGCAAAAACTGTTAAGCCACCCA
CTGACCCAGAGCTTGCAGCAATTCGAGTTAACAAAATTCTGCCAATTCTCCAAGATTTACAAAGTGCAGA
CCAGTCAAAGAGATCAACTGCTGCAACTGCCATTGCGAACCTCGTTGACGATACAAAATGTCGAAAGTTA
TTCTTGAGAGAGCAAATTGTTCGTATTCTACTCGAACAAACCCTTACAGACTCAAGCATGGAAACTAGAA
>lcl|NW_001820817.1_gene_205 [locus_tag=SS1G_12233] [db_xref=GeneID:5483157] [partial=5',3'] [location=complement(<502136..>503461)] [gbkey=Gene]
ATGATCTGTAATACGCTCGGTGTTCCACCCTGCAACAGAATTCTTAAGAAATTCTCCGTTGGCGAGAGTC
GTCTCGAAATTCAAGACTCAGTACGAGGCAAAGATGTCTACATCATTCAATCGGGTGGAGGAAAGGCCAA
TGATCACTTCGTGGATCTTTGCATTATGATCTCCGCATGCAAAACTGGCTCTGCCAAGCGCGTCACTGTC
GTCCTTCCTTTGTTTCCTTATTCACGACAACCTGATCTGCCATACAACAAGATTGGCGCACCACTTGCCA
>lcl|NW_001820834.1_gene_1034 [locus_tag=SS1G_02099] [db_xref=GeneID:5493612] [partial=5',3'] [location=<2692251..>2693298] [gbkey=Gene]
ATGGCTTCTGTTTACAAGTCATTATCAAAGACCTCTGGTCATAAAGAAGAAACCCCGACTGGTGTCAAGA
AAAACAAGCAAAGAGTTTTGATCTTGTCTTCAAGAGGAATAACTTACAGGTATATAAATTTGTACCGATG
CGATGCAAAAAATCGCAGGAAAATGCTAACTCTACAACTTAGACATCGACATCTCCTCAATGACCTTGCG
TCCCTACTTCCCCACGGTAGGAAAGATGCGAAACTCGATACCAAGTCAAAGCTTTATCAATTGAATGAAT
>lcl|NW_001820830.1_gene_400 [locus_tag=SS1G_05227] [db_xref=GeneID:5489764] [partial=5',3'] [location=complement(<1032740..>1033620)] [gbkey=Gene]
ATGGCGGACGGATGTAAGTTAATTGATGTTCCTACTATTCCAGACTAATATTTGTTCTCGTCCCTACAAT
GCATTCGGAACGGATGGTACTCAGTTAACTTTGTAACTAATACAACGTCTAGTAAATGACCAAAGAACTG

I am new in python so I tried to come up with something like this:

results = 
f = open("test.txt", 'r')

while True:
 line = f.readline()
 if not line:
 break
 file_name = line.split("locus_tag")[-1].strip()
 f.readline() # skip line 
 data_seq1 = f.readline().strip()
 f.readline() 
 data_seq2 = f.readline().strip()
 results.append((file_name, data_seq1))

I think, the most trival way to solve your issue is by using regex like this example:

import re

results = 
# Open the file in the 'read' mode
# with statement will take care to close the file
with open('YOUR_FILE_PATH', 'r') as f_file:
 # Read the entire file as a one string
 data = f_file.read()
 # Here we search for the string that begins with '>lcl'
 # and in which we find the [locus_tag=...] and [localtion=...]
 results = re.findall(r'>lcl.*[locus_tag=(.*?)].*[location=(.*?)]', data)

for locus, location in results:
 print(locus, location)

Output:

SS1G_08319 <504653..>506706
SS1G_12233 complement(<502136..>503461)
SS1G_02099 <2692251..>2693298
SS1G_05227 complement(<1032740..>1033620)

Another variation using a dict as a result and by splitting lines:

import re

results = 
with open('fichier1', 'r') as f_file:
 # Here we split the file's lines into a list
 data = f_file.readlines()
 for line in data:
 # Here we search for the lines that begins by '>lcl'
 # and same as the first attempt
 results.update(re.findall(r'^>lcl.*[locus_tag=(.*?)].*[location=(.*?)]', line))

for locus, location in results.items():
 print(locus, location)

Edit: Creating a DataFrame and exporting it into a csv file:

import re
from pandas import DataFrame as df

results = 
with open('fichier1', 'r') as f_file:
 data = f_file.readlines()
 for line in data:
 results.update(re.findall(
 r'^>lcl.*[locus_tag=(.*?)].*[location=(.*?)]',
 line
 ))

df_ = df(
 list(results.items()),
 index=range(1, len(results) + 1),
 columns=['locus', 'location']
)
print(df_)
df_.to_csv('results.csv', sep=',')

It will prints and creates a file called results.csv:

 locus location
1 SS1G_12233 complement(<502136..>503461)
2 SS1G_08319 <504653..>506706
3 SS1G_05227 complement(<1032740..>1033620)
4 SS1G_02099 <2692251..>2693298

I would like to present two alternative solutions. One that will extract any set of named tags on your line using regular expressions, and another which is a complete travesty but shows a way to do it without regular expressions.

Generic Regex Solution

import re

def get_tags(filename, tags, prefix='>lcl'):
 tags = set(tags)
 pattern = re.compile(r'[(.+?)=(.+?)]')

 def parse_line(line):
 return m.group(1): m.group(2) for m in pattern.finditer(line) if m.group(1) in tags

 with open(filename) as f:
 return [parse_line(line) for line in f if prefix is None or line.startswith(prefix)]

This function returns a list of dictionaries keyed by the tags you are interested in you would use it like this:

tags = ['locus_tag', 'location']
result = get_tags('test.txt', tags)

You could use the result to get the exact printout you want:

for line in get_tags('test.txt', tags):
 print(*(line[tag] for tag in tags))

This has the advantage that you can use the results as you chose later, and configure which tags you extract.

No Regex Solution

This version is just something I wrote to show that is possible. Please do not emulate it, as the code is a pointless maintenance burden.

def get_tags2(filename, tags, prefix='>lcl'):
 tags = set(tags)

 def parse_line(line):
 items = [tag.split(']')[0].split('=') for tag in line.split('[')[1:]]
 return dict(tag for tag in items if tag[0] in tags)

 with open(filename) as f:
 return [parse_line(line) for line in data if prefix is None or line.startswith(prefix)]

This function behaves just like the first one but the parsing function is a hot mess by comparison. It's also much less robust, e.g. because it's assumed that all your square brackets are more or less matching.

Here is an IDEOne link showing off both methods: https://ideone.com/X2LKqL