fasta file parsing for blast module

[msarrias]

Hi,

For the blast module, I think it would be convenient to modify the way fasta files are read, so that it is not sensitive to the number of lines in the file.

---

update:

As I understand it, a sequence record in a Fasta format consists of a single line of description starting with >, followed by a line(s) of sequence data, i.e. the sequence can be split into multiple lines and does not necessarily have to be a single line (as required by gget).

For example, if the sequence record is as follows:

>ENSFALP00000003283
MVKAAKEIQMENPREAESPSTGATCSPPEEQAKKPSMLCFKKRKKSCKKELTVKDACEGAS
EEKSQCVSADQEEAKASNPSQSSQKGTWAAIKNLARPRKRQKSCSRKKVPSDSQVQLEVD
AEESCAQELPKKRASSGVKMPCVRFSRGKKKPSPSEAVEESEGSVQANEVMGVVNKASEE

gget will raise an exception since the sequence is split into 3 lines.

[lauraluebbert]

Hi Marina,

Thank you for your suggestion! :) The reason gget blast currently only takes into account the first sequence is that each sequence will have many rows of output. To BLAST all sequences in a fasta, I would have to concatenate the outputs while making clear which results belong to which sequence.

For now, I would recommend passing each sequence in a fasta to gget blast in a loop. For example, the code below will pass each sequence in "test.fa" to gget blast and then save the results in a csv named after the sequence ID (text next to the ">").

from Bio import SeqIO
import gget

for record in SeqIO.parse("test.fa", "fasta"):
    df = gget.blast(record.seq)
    df.to_csv(record.id)

[msarrias]

Hi Laura,

Thanks for your reply. I don't think I explained myself correctly, sorry for that. I have included an update in the discussion.

[lauraluebbert]

I understand; thank you for clarifying! I will take a look at it