Alignment & BLAST: what is this sequence?

Imagine someone hands you a string of letters like GATTACA and asks, what is this? In this lesson you will learn how to answer that by comparing your sequence to ones that scientists have already identified. The two core tools for doing this are sequence alignment and a search program called BLAST.

What a sequence is

A DNA sequence is written with just four letters, A, C, G, and T, where each letter stands for one chemical building block called a nucleotide. A protein sequence uses 20 different letters instead, one for each building block called an amino acid. Either way, a sequence is simply that string of letters, read from left to right.

Aligning two sequences

Aligning means sliding two sequences side by side so that matching letters line up in columns, scoring a point for each match and a penalty for each mismatch or gap. A gap, shown as a dash, marks a spot where one sequence has an extra letter that the other is missing, representing an insertion or a deletion; percent identity is simply the share of aligned columns that match. Here is a short alignment of two DNA pieces, with vertical bars marking the matches:

Query  1  ACGTACGTA  9
          ||| ||| |
Sbjct  1  ACGAACG-A  8

Identities = 7/9 (78%), Gaps = 1/9 (11%)

Searching with BLAST

BLAST stands for Basic Local Alignment Search Tool. You give it one sequence, called your query, and it rapidly aligns that query against a huge database of known sequences, then returns a ranked list of the closest matches, called hits. For each hit you mainly read three numbers: percent identity, query coverage, and the E-value.

Percent identity tells you how alike the matched regions are, and query coverage tells you what fraction of your query actually took part in the alignment, so a strong hit usually scores high on both. The E-value then estimates how likely a match this good could have appeared purely by chance.

To run this yourself, open the NCBI BLAST website, paste your sequence into the query box, and choose the right program: blastn compares DNA against DNA, while blastp compares protein against protein. Click the BLAST button, wait a moment, and read the results table, where hits are listed best first, usually sorted by E-value. The top hit's description often tells you what gene, protein, or organism your sequence comes from.