BLAST Annotating Sub-sequences of Features with Not Normalized Identity Scores

Question

[Fadi Bakoura]

I'm using BLAST annotation to annotate protein features using the online tool

Below is a screenshot of the entered queries.

Below is a screenshot of the result.

My question is there is a parameter to let BLAST normalizes the identity score for the 12his feature? It reports 100% where in fact it's only 50%

Answer 1

[massa.kassa.sc3na]

The percent identity relates to the alignment. The alignment in BLAST is LOCAL, thus only part of the query and subject may be aligned in the HSP. See the alignemnts for yourself by clicking the "Alignments" to see that the two aligned sequence portions are 100% identical.

The "query coverage" tells you that only part of the query is aligned and is should be first signal for you that this might happen.

For the second part, of your question, I'm not sure if there is a parameter in the BLAST itself, because it sounds like that you are trying to do global alignment. The global will score the two sequences end-to-end. Notorious example of such algorithm is Needleman-Wunsch.