Entering edit mode
8.0 years ago
PJ.Tatlow
•
0
So I'm trying to analyze the starts of multiple genes across a group of different bacteriophages to make sure we are annotating them consistently.
I'm using a local blast database to see if the starts align 1-1, but something weird is happening. I have the following two sequences:
>vB_EamM_Stratton_8
MKKLLVGLLAVLAFSAHAADVTITTGQQGLTYNATYGVNLASALNEFGNKSTVVPSKGSLDNLDKVAAGSAQIGFTQADAFQFWRSQHPNEAQNVDIVGQLGDECVFVAVKKEGKVGGEADLKEGIKIAVGEPDSGSYASWQYLQGLVKQYAKAETYAKGGIRSLSKVATGEYDAFMWVSAPGKNNKFLEAVNQDSSGLKLIDMNNWNVNDKLPNGQAVYEKKSAKVDNSAWFGGDVKVPCTKTLVVANTDAGDELLEAVSTILLKNAPRIMGTGK
>vB_EamM_Huxley_9
MRKVLFGILAALTFSMSAVAADEVVITTGQQGLTYNSVYGVNLAGAISEFGTRARVIPSKGSLDNLDKVAAGEAQIGFTQADAFQFWRAKHPNEAQNVDIVGQLARECVWVAVKKDGKVKDAKDLTDGVKIAVGEPDSGSYASWQYLQQLVKEYSKAETYAKGGIRSLSKVATGEYDAFLWVSAPDKSNKFLEAVNQDSSGLELIDMSTWNVNDKLPNGQSVYTKETAKTDKTAFFGKKVDVPCTTTLVVANTDAGDDLLETVSTILLKNNSRVMGDKK
When I blast Stratton_8 against the database I get that it is a 1-1 with Huxley_9. But when I blast Huxley_9, it's a 16-14 with Stratton 8.
Any ideas what I can do to get consistent results from blast?
Thanks!
Since blast tries to find the best local alignment it may be finding that partial match better than the full sequence. Take a look at this thread and consider using the -qcov_hsp_perc 100 option if you only want to see hits which include full length query sequence.
Thanks! I'll give that a shot.
As far as I have understood you aligned protein sequences.
The sequences you showed do not have ">"-sign just before their headers.
It's incorrect, try to change it everywhere. They are not fasta-files without ">"-sign.
Blast may not understand where your sequences actually start.
How have you made your database? With makeblastdb or in a different way?
What program did you use? Was it blastp in comand line or something different?
I had the > in the fasta file, but it tried to format it as a quote when I typed it in online so I just got rid of it! But don't worry, it is there in the file. I used makeblastdb to make the database, and I used blastp to make the queries.