• Biotechnology: bioinformatics, biological data and the computer

What is BLAST?

Written with AI (How we create content)Target level: Grade 11Last updated: Report an error
Quick answer

BLAST, the Basic Local Alignment Search Tool, is a free program from the NCBI that compares a DNA or protein sequence, the query, with millions of sequences in a database and lists the most similar ones within seconds. Each hit comes with its percent identity, how much of the query it covers and an E-value, the number of equally good matches expected by chance.

The topic explained

A search engine for sequences

What is BLAST, and why does nearly every biotechnology project use it? A researcher who has just read a new sequence wants to know what it resembles. Aligning it exactly with each of millions of database sequences, with the Smith–Waterman method, would take far too long. BLAST, the Basic Local Alignment Search Tool, published in 1990 by Stephen Altschul, Warren Gish, Webb Miller, Eugene Myers and David Lipman, finds almost the same answers in seconds. Lipman later directed the NCBI, which runs BLAST free on the web.

The sequence you search with is the query; the sequences found are hits. BLAST is local: it finds the regions that match well and does not force the whole query to align. A common misreading is that BLAST says what a sequence is. It says which known sequences it resembles; the conclusion is the researcher's.

A query bar above a database sequence bar, joined by a short highlighted seed match in the middle, with arrows pointing outward from the seed in both directions to show it being extended.seed matchquerydatabase sequenceextend
Schematic of seed and extend, the idea that makes BLAST fast.

Seed and extend

BLAST is fast because it does not try every alignment. It cuts the query into short words of a few letters and looks up where those words, or close variants, occur in the database, which is indexed in advance like a dictionary. Each place where a word matches is a seed. BLAST then extends each seed letter by letter in both directions as long as the score keeps rising, and keeps the high-scoring segment pairs. Regions with no seed are never examined, which is why BLAST can, rarely, miss a weak match an exact method would find.

There are several programs. blastn compares a DNA query with DNA, blastp a protein with proteins, and blastx translates a DNA query in all six reading frames and compares the results with proteins, useful for an unknown gene. Protein searches are more sensitive for distant relatives, because the genetic code hides many DNA changes that leave the protein unchanged.

Five steps from top to bottom: split the query into short words; find word hits in the database; extend the hits both ways; keep the high-scoring pairs; rank the hits by E-value.1. Split query into words2. Find word hits3. Extend hits both ways4. Keep high-scoring pairs5. Rank hits by E-value
BLAST checks only the regions around word hits, not every possible alignment.

Reading the table of hits

Hits are listed with their percent identity, their query cover, the share of the query that aligns, and their E-value. The E-value is the number of hits with this score or better that a search of this database would be expected to find by chance. An E-value of 1 means one such hit is expected by chance, so it is meaningless; 3e-12, that is 3×10−123 \times 10^{-12}, means a chance hit is extremely unlikely. Because a bigger database offers more chances, the same score in a database ten times larger gets an E-value ten times larger.

In the invented table, hit 1, 98% identical over the whole query with E-value 2e-50, is almost certainly the same gene or a very close relative. Hit 2, 62% over 80% with 3e-12, is probably a related gene. Hit 3, 31% over 20% with E-value 4.5, is no evidence of anything. Students often judge by identity alone; a short stretch can be 100% identical by chance, so cover and E-value are read too.

A table of three invented hits: hit 1 with 98% identity over the whole query and E-value 2e-50; hit 2 with 62% identity over 80% and E-value 3e-12; hit 3 with 31% over 20% and E-value 4.5.HitIdentityQuery coverE-valuehit 198%100%2e-50hit 262%80%3e-12hit 331%20%4.5
Invented results. 2e-50 means 2 × 10⁻⁵⁰.

What BLAST cannot tell

A strong hit is only as good as the record it points to. Database entries can carry wrong names or functions copied from earlier entries, and a tiny E-value shows that the similarity is not chance, not that the two sequences do the same job. Researchers therefore check several hits, look at curated entries, and test the conclusion in the laboratory. Used this way, BLAST is often the first step of a project: it names the likely gene, points to relatives whose function is known, and supplies the sequences from which primers and probes are designed.

Free printable worksheet

Practice: what BLAST is

For 11th grade biotechnology students. Use the figures and table on the sheet, then check your answers.

Free, no sign-up. The answer key prints on the second page. Regenerating a new worksheet on the same topic needs a free account.

Free printable worksheetlessoons.com
Practice: what BLAST isName:
  1. What does BLAST stand for, and what does it do?

  2. Name the short highlighted match marked with a question mark, and say what BLAST does next with it.

    The seed-and-extend schematic with the name of the short highlighted match between the two sequences replaced by a question mark.?querydatabase sequenceextend
  3. Fill in step 3 and explain why this approach is faster than aligning the query with every sequence in full.

    The five steps of a BLAST search with step 3 replaced by a question mark.1. Split query into words2. Find word hits3. ?4. Keep high-scoring pairs5. Rank hits by E-value
  4. Which program would you use for each: a DNA query against DNA; a protein against proteins; an unknown DNA sequence that may code for a protein?

  5. Interpret each of the three hits in the table.

    The table of three invented hits with their identity, query cover and E-value.HitIdentityQuery coverE-valuehit 198%100%2e-50hit 262%80%3e-12hit 331%20%4.5
  6. A hit has E-value 0.002 in one database. What E-value would the same score get in a database ten times larger?

  7. A friend says the top BLAST hit proves her sequence is an enzyme for breaking down starch. Give two reasons to be careful.

More topics and worksheets at lessoons.com© 2026 Lessons

Common questions

What is a good E-value in BLAST?

The lower the better. Values far below 0.001 usually mean the match is not chance; values near 1 or above mean such a match is expected by chance. The cut-off depends on the question and the database size.

What is the difference between blastn and blastp?

blastn compares nucleotide sequences, DNA or RNA, with a nucleotide database. blastp compares a protein sequence with a protein database, which is better for finding distant relatives.

Is BLAST free?

Yes. The NCBI runs BLAST free on its website, and the program can also be downloaded to run on one's own computer and databases.

What is query cover?

The percent of the query sequence that is included in the alignment with a hit. A high identity over a low cover means only a short part of the query matches.

Can BLAST miss a real match?

Yes, occasionally. Because it starts only from seed words, a very weak or scattered similarity with no seed can be missed, which slower exact methods might find.

Still have a question? Ask Lessi

Any question about this subject. Lessi answers right here, step by step.

Try asking

Asking needs a free account; your question is kept while you sign up. Answers are written by AI and can contain mistakes. Lessi doesn’t give medical or legal advice.

Learn it with a teacher

Learn this topic live with Lessi

Lessi, the AI teacher, explains it out loud on a board, with short quizzes, at your pace. The first 60 minutes are free.