PyFasta - Splitting large headers

[quacksawbones]

I am receiving some strange issues with PyFasta concerning splitting large headers within a Python script.

This is the code:

from pyfasta import Fasta

compare = Fasta('./test_pyfasta.fasta', key_fn=lambda key: key.split()[0])
print compare.keys()

when using the two sequences in my test dataset:

Ch_kdna_pair1_1|Len:90
AAAGAAGTTATCACTGTGTTTATACGTAAGTTATCAGGCCTGCTACACGTAAAGTATCAGACTATTAAACAAGTCACCC
CGGGGTCAAAG
Bx_kdna_pair1_4 Len:90
AAGCTGCGTCCTGACCGCGAAAAGACTTAGGCCCCGTTTGTTAGCTTAGTATTTTTAAAGTATTCTGAGAATACTACAT
TTTTTCAGATT

it will only split the space, NOT the pipe. I have some other files where I intend to (hopefully) split by "." rather than by " " or "|". Is there any way to specify the delimiting character, similar to the string.split function?

[brentp]

key_fn is a function, so you can set it however, you like:

Fasta('t.fasta', key_fn=lambda key: key.split("|")[0]).keys()

or you can even use a regexp to split on either space or pipe:

Fasta('t.fasta', key_fn=lambda key: re.split("[\s\|]", key)[0]).keys()

please let me know if this addresses your use-case.

[quacksawbones]

Hi Brent,

Thanks for getting back to me so quickly. Unfortunately, putting both snippets above into my code (the only change is replacing t.fasta with my file location) returns:

Traceback (most recent call last):
  File "test_pyfasta.py", line 12, in <module>
    print compare.keys()
AttributeError: 'list' object has no attribute 'keys'

Irrespective of the characters I use. While I am using the argparse module, I get exactly the same result whether I directly access the file or use an argument in the command line. The only modules I have imported are "argparse" and "from pyfasta import Fasta".

Thanks for your help Brent.

[brentp]

note that my snippet already called keys. so you dont need to call it again.
On Jan 2, 2012 10:00 PM, "quacksawbones" <
reply@reply.github.com>
wrote:

Hi Brent,

Thanks for getting back to me so quickly. Unfortunately, putting both
snippets above into my code (the only change is replacing t.fasta with my
file location) returns:

Traceback (most recent call last):
 File "test_pyfasta.py", line 12, in <module>
   print compare.keys()
AttributeError: 'list' object has no attribute 'keys'

Irrespective of the characters I use. While I am using the argparse
module, I get exactly the same result whether I directly access the file or
use an argument in the command line. The only modules I have imported are
"argparse" and "from pyfasta import Fasta".

Thanks for your help Brent.

---

Reply to this email directly or view it on GitHub:
#6 (comment)

[quacksawbones]

Hi Brent,

Unfortunately, it still does not want to play. I have tried it with Python 2.5 and 2.7, x86 and x64 versions and it will not split them using the pipe (or any othe character apart from space.

Any suggestions? I have been using ActivePython with NumPy and SciPi as well as PyFASTA installed, but I have tested it with the standard version of Python (from www.python.org) with no other customisations to the distributions (except NumPy, SciPy and PyFasta) and it still does not work.

What version of Python are you using and what modules have you got installed?

Thanks again Brent.

[brentp]

@quacksawbones please paste the full, exact python code, and the fasta into the ticket or a gist.

[brentp]

you might try deleting the test_pyfasta.fasta.gdx file then re-running.

[quacksawbones]

Hi Brent,

Bingo! The *.gdx file was causing the issue. It is now working as required on Python 2.7 x64.

Thanks for your assistance (and, more importantly, patience!)