是否有代码可以从带有 ID 列表的大型 sdf 文件中提取完整的分子记录？

Question

I am using Phyton 3.7 in Spider

I try to extract complete molecular records from a large sdf file with an small list of IDs listed in a txt file and writing them into one new sdf file.

More specifically, I have a selected list of about 500 chemical molecule IDs in lines, one ID per line (ten number digits each ID) whose molecular details are contained into a large sdf file of about 2G (300000 molecules, each record contains about 400 lines of code between their ID to the final $$$$ line)

I need to extract the complete 500 records of the IDs into a single sdf file from the large sdf 2G file for further studies.

I tried the somehow similar and partial python scripts from the stackoverflow and google but not a single one work? Could anyone give a hint or a few lines of code to test?

Thank you julio

As suggested (thank you Andrej: great idea), to simplify the problem, I designed small samples of the files. Each line is separated by \n in the originals. I added the positional information to each record to facilitate follow up of results. f1.txt contains 3 IDs f2.sdf contains a simplifyed sample of the large 2G data base f3.sdf contains the desired file, in this case, for the 3 IDs

f1.txt

SN00061212
SN00134795
SN00107686

f2.sdf

SN00039109
 MOLSOFT 05232012283D, 1 in the large sdf list

about 400 lines more of code

$$$$
SN00357061
 MOLSOFT 05232012283D, 2 in the large sdf list, 

about 400 lines more of code

$$$$
SN00134795
 MOLSOFT 05232012283D, 3 in the large sdf list

about 400 lines more of code

   $$$$
SN00061212
 MOLSOFT 05232012283D, 4 in the large sdf list

about 400 lines more of code, one in the short txt list SN1

  $$$$
SN00134796
 MOLSOFT 05232012283D, 5 in the large sdf list

about 400 lines more of code

  $$$$
SN00134795
 MOLSOFT 05232012283D, 6 in the large sdf list

about 400 lines more of code, one in the short txt list SN2

  $$$$
SN00333333
 MOLSOFT 05232012283D, 7 in the large sdf list

about 400 lines more of code

  $$$$
SN00145791
  MOLSOFT 05232012283D, 8 in the large sdf list

about 400 lines more of code

  $$$$
SN00107686
 MOLSOFT 05232012283D, 9 in the large sdf list

about 400 lines more of code, one in the short txt list SN3

$$$$ 

---

f3.sdf

SN00061212
 MOLSOFT 05232012283D, 4 in the large sdf list

about 400 lines more of code, one in the short txt list SN1

  $$$$
SN00134795
 MOLSOFT 05232012283D, 6 in the large sdf list

about 400 lines more of code, one in the short txt list SN2

  $$$$
SN00107686
 MOLSOFT 05232012283D, 9 in the large sdf list

about 400 lines more of code, one in the short txt list SN3

$$$$

Answer 1

You can use re module for the task:

If f1.txt contains:

SN00061212
SN00134795
SN00107686

f2.sdf contains:

SN00039109
 MOLSOFT 05232012283D

about 400 lines more of code

$$$$
SN00357061
 MOLSOFT 05232012283D

about 400 lines more of code

$$$$
SN00061212
 MOLSOFT 05232012283D

about 400 lines more of code, one in the short txt list SN1

  $$$$
SN00134796
 MOLSOFT 05232012283D

about 400 lines more of code

  $$$$
SN00134795
 MOLSOFT 05232012283D

about 400 lines more of code, one in the short txt list SN2

  $$$$
SN00333333
 MOLSOFT 05232012283D

about 400 lines more of code

  $$$$
SN00145791
  MOLSOFT 05232012283D

about 400 lines more of code

  $$$$
SN00107686
 MOLSOFT 05232012283D

about 400 lines more of code, one in the short txt list SN3

$$$$

Then this script:

import re

with open('f1.txt', 'r') as f_in:
    desired_ids = set(line.strip() for line in f_in if line.strip())

expr = r'({}.*?^\s*\$\$\$\$)'.format(r'^\s*(?:' + r'|'.join(re.escape(i) for i in desired_ids) + r')')
r = re.compile(expr, flags=re.DOTALL|re.M)

with open('f2.sdf', 'r') as f_in, open('f3.sdf', 'w') as f_out:
    for m in r.finditer(f_in.read()):
        print(m.group(0), file=f_out)

Produces f3.sdf :

SN00061212
 MOLSOFT 05232012283D

about 400 lines more of code, one in the short txt list SN1

  $$$$
SN00134795
 MOLSOFT 05232012283D

about 400 lines more of code, one in the short txt list SN2

  $$$$
SN00107686
 MOLSOFT 05232012283D

about 400 lines more of code, one in the short txt list SN3

$$$$

EDIT:

You can see the regex live on regex101

The re.DOTALL means that the dot . character matches newlines also. The re.M (or re.MULTILINE ) means, that ^ character will match beginning of the line, not just beginning of the file. More in official re documentation .