You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I use blast to find DNA parts in my lab's inventory, but today I came across something strange when using blasterjs.
If a given subject sequence produces more than one alignment with the query, all alignments are coloured the same at the "blast-multiple-alignments" section.
A scenario where this could be a problem is when looking for a C-terminal fusion instead of an N-terminal fusion. The displayed result would be the same, even though the BLAST result is not.
Conceptually, if the subject contains the same parts as the query, but ordered differently, the displayed alignments are exactly the same as if the subject had been identical. Coverage is 100% in both cases and, from looking at the plots, you would not realise that query and subject are very different.
It would make sense to include a thin bounding box around each rectangle (or any other suitable visual element) to show clearly when a subject has more than one alignment with the query.
Also a hint to the strand sense of the alignments would be informative.
I have more ideas and comments, which I'll share if you'd like me to.
Best,
N
The text was updated successfully, but these errors were encountered:
Hi, great work!
I use blast to find DNA parts in my lab's inventory, but today I came across something strange when using blasterjs.
If a given subject sequence produces more than one alignment with the query, all alignments are coloured the same at the "blast-multiple-alignments" section.
A scenario where this could be a problem is when looking for a C-terminal fusion instead of an N-terminal fusion. The displayed result would be the same, even though the BLAST result is not.
Conceptually, if the subject contains the same parts as the query, but ordered differently, the displayed alignments are exactly the same as if the subject had been identical. Coverage is 100% in both cases and, from looking at the plots, you would not realise that query and subject are very different.
It would make sense to include a thin bounding box around each rectangle (or any other suitable visual element) to show clearly when a subject has more than one alignment with the query.
Also a hint to the strand sense of the alignments would be informative.
I have more ideas and comments, which I'll share if you'd like me to.
Best,
N
The text was updated successfully, but these errors were encountered: