BioPython local blast different results than blast+ local blast

Question

[s.l.rinzema]

Hi quick question!

I've done the same blast (tblastn) twice, the first one with the blast+ module and the second one with the standalone NCBI blast+ suite. Which uses the same blast module according to the cookbook. Though both give different results and as far as I can see the parameters are the same. Anybody know why? Am I missing something?

Thanks in advance!

[Sej Modha]

If you mean standalone BLAST by

the blast+ module

then you might want to check the version of local BLAST databases as using the different versions of the BLAST DBs would give you different output.

[s.l.rinzema]

Thanks for the quick reply,

I forgot to mention I blast fasta against fasta. Not against a database.

[st.ph.n]

You can make any fasta into a blast database with makeblastdb

[s.l.rinzema]

Thank you, but I don't think that's helpful and it doesn't explain why I get two different results.

[Sej Modha]

Is it possible to provide an example of a query and the corresponding output e.g. in the BLAST tabular format that would help clarify what you mean by different results?

[s.l.rinzema]

Hi I accidentally replied to you in another answer

[s.l.rinzema]

Ofcourse! Here is my standalone BLAST commandline:

tblastn -query fasta1.fna -subject fasta2.faa -out outputfile.txt

the python code i use to from biopython:

from Bio.Blast.Applications import NcbitblastnCommandline
cline = NcbitblastnCommandline(query=QUERY, subject=SUBJECT, outfmt=5, out=out_file)
cline()

I now see that my BioPython code's output is in XML format and my commandline is in pairwise (the standard). If i add the "-outfmt 5" to the commandline I get the same result as my BioPython code but it's different as with the pairwise. Now my question is why? Still thank you! I guess this solved my problem and brought me to another

[Sej Modha]

Maybe top BLAST hits were not in the same order? You could get output in the tabular BLAST format using -outfmt 6 option and parse the hits for each query and check if it is just the order that was different or there were discrepancies in the corresponding evalue and bitscore.

[s.l.rinzema]

Thank you for your help, I already fixed it by changing the output to XML. With both outputs as XML the evalues and hits were exactly the same!