evalue

An E-value, or e-value, is a statistical measure used in bioinformatics to assess the significance of sequence alignments. It represents the expected number of database matches with a score as good as or better than the observed one that would occur by chance in a database search of a given size.

E-values are derived from the scoring system used for the alignment and typically rely on Karlin-Altschul statistics. A common form is E = K m n e^{-λS}, where S is the alignment score, m and n are the lengths of the query and database, and λ and K are constants determined by the scoring matrix and gap penalties. The constants reflect the statistical properties of the scoring scheme and the search space.

Interpretation and usage: a small E-value indicates a more significant match. However, the E-value is not a direct probability. It is related to the likelihood of finding such a match by chance in the specified database. The corresponding probability can be estimated as p = 1 − e^{-E} in this framework, though practical interpretation often focuses on threshold values.

In practice, E-values are widely used in tools such as BLAST, FASTA, and PSI-BLAST to report results of protein or nucleotide similarity searches. They depend on the size and composition of the database, the scoring matrix, and gap penalties. Users should consider E-value thresholds in context, and complement them with additional evidence and biological relevance rather than relying on a single metric.