1
$\begingroup$

Using the BLASTn webtool with,

Query Seq: CAATTGCGAAGCCACATTATT

Subject Seq: CP031332.1

BLASTn returns a single hit to CP031332.1 with query coverage at 100%.

However if one looks at the alignment, the query coverage is clearly 17/21 ≈ 81%. Could someone explain to me the reason for this discrepancy?

Edit,

Web version: https://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastn&PAGE_TYPE=BlastSearch&LINK_LOC=blasthome

Output: https://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Get&RID=RCK3F55Y114

$\endgroup$
0
1
$\begingroup$

To expand on the excellent answer by haci:
BLAST reports (a) the summary of alignments and (b) individual alignments, see below. Note that your link to the results expired (as NCBI BLAST results links tend to do). It is considered best practice to post the text version of the relevant portion of the results in the question itself.

The summary provides an overview of all the query sequence individual hits in the subject sequence. Because the individual hits (alignments, or ranges) together cover 100% of the query and each of these has the max identity = 100%, that is what it reports in the summary. It is not very intuitive for many users.

You should always look at the individual hits in detail. The above 100% coverage in the summary consists of ranges 1 and 14, shown below. The top hit (range 1) covers only 17 out of 21 nt with 100% identity, and this is the most relevant information from BLAST. The other hit (range 14) covers 10 out of 21 nt with 100% identity. This is the cause of the discrepancy.

These 2 alignments are not close together in the subject sequence. Depending on your goal, choose one or more of the individual alignments and consider alignment length and % identity of those hits individually, rather than summarized together as in the BLAST report summary. For most applications, the top hit (range 1) would be the most relevant alignment.

Program: BLASTN 
Subject:Neisseria meningitidis strain M22814 chromosome, complete genome ID: CP031332.1(nucleic acid) Length: 2190201
Query #1: seq1 Query ID: lcl|Query_56011 Length: 21

Sequences producing significant alignments:
                                                                  Max    Total Query   E   Per.                  
Description                                                       Score  Score cover Value Ident  Accession        
Neisseria meningitidis strain M22814 chromosome, complete genome  34.2   57729 100%  0.001 100.00 CP031332.1       


Alignments:
>Neisseria meningitidis strain M22814 chromosome, complete genome
Sequence ID: CP031332.1 Length: 2190201 
Range 1: 2074883 to 2074899

Score:34.2 bits(17), Expect:0.001, 
Identities:17/17(100%),  Gaps:0/17(0%), Strand: Plus/Plus

Query  1        CAATTGCGAAGCCACAT  17
                |||||||||||||||||
Sbjct  2074883  CAATTGCGAAGCCACAT  2074899

...

Range 14: 354292 to 354301

Score:20.3 bits(10), Expect:15, 
Identities:10/10(100%),  Gaps:0/10(0%), Strand: Plus/Minus

Query  12      CCACATTATT  21
               ||||||||||
Sbjct  354301  CCACATTATT  354292

$\endgroup$
0
$\begingroup$

Your blast result is not composed of a single alignment of all of the length of your query vs the subject but from a series of alignments: in each a subsequence of your query is used and these alignments are sorted by "E value", "score", "percent identity", ...

Now in your case, you are talking about the first hit, in which a subsequence of your query (positions 1-17) match perfectly to a portion of the subject. Query coverage is not calculated from this first hit though, it takes all of the alignments into account and again in your case all 21 bases were used in one or more alignments. I believe it would be fair to filter out all but the first hit given the E-values and re-calculate the query coverage as 17/21.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.