BLAST Searching
BLAST (Basic Local Alignment Search Tool) takes your sequence, the query, and searches a large database for sequences that resemble it. It is the standard way to answer "what is this sequence, and what is it related to?", for instance for an unknown gene, a sequenced fragment or a protein of interest.
How it works
Comparing a query against billions of letters with a full alignment would be far too slow. BLAST first finds short exact or near-exact "words" shared with database sequences, then extends those seeds into longer local alignments and scores them. Scores are converted to bit scores and E-values that account for the size of the database, so results can be compared between searches.
Which program
A nucleotide query against nucleotide sequences is blastn; a protein query against proteins is blastp. Protein comparison finds far more distant relatives than DNA comparison, because the genetic code is redundant and amino acids tolerate similar swaps, so for a coding sequence a protein search is often more sensitive.
Interpreting the top hit
The best hit is not automatically the right answer. A tiny region can match perfectly with an excellent E-value while the rest of the query matches nothing, so check coverage. Several hits in the same species or gene family strengthen a conclusion; a lone hit does not. Low-complexity or repetitive regions (poly-A, simple repeats) and cloning-vector sequence can produce misleading hits. Short queries, under about 30 bases, give weak statistics. A high identity to a distant species with only partial coverage is a different finding from 100% over the full length.
How Aaron's Archive does it
| Item | What the tool does |
|---|---|
| Where the search runs | DNA and RNA go to NCBI's BLAST service (searching its core nucleotide collection); protein searches go to EMBL-EBI, with NCBI as a fallback. This is not computed on this site's own server, so your sequence is sent to that service (see the privacy policy). |
| Waiting | A search is queued remotely and usually takes from a few seconds to a few minutes; the page checks progress for you. |
| Result | The best matches with E-value, identity, coverage and the aligned region. |
Pitfalls
Do not send confidential or unpublished sequences to any public BLAST service. If a search fails the page says whether the problem is on your side, NCBI's or EMBL-EBI's, and what to try next.
Common questions about BLAST
What does the E-value mean?
The number of hits with a score at least this good that you would expect by chance in a database of this size. Smaller is better: 0.001 is a chance match about once in a thousand searches.
What is a good E-value?
There is no universal cut-off. Values below about 1e-5 are often used to suggest a real relationship, and much smaller values are common for close relatives. Also look at identity and how much of your query is covered.
What is the difference between blastn and blastp?
blastn compares a nucleotide query with nucleotide sequences; blastp compares a protein query with proteins. Protein searches find more distant relatives, because protein sequence changes more slowly than DNA.
How is BLAST different from aligning two sequences?
BLAST searches a whole database for matches to one query. To compare two sequences you already have, use the pairwise alignment tool.