BLOSUM62

BLOSUM62 is a protein sequence substitution matrix used to score alignments between protein sequences. It is part of the BLOSUM family developed by Stephen Henikoff and colleagues in 1992. The name BLOSUM stands for Blocks Substitution Matrix. The 62 in the name refers to the maximum percent identity used to construct the matrix: the blocks from which substitution frequencies were estimated contain no more than 62% identical residues.

Scores are log-odds measures of the observed substitution frequency versus background amino acid frequencies, expressed in bits. The matrix comprises 20x20 entries for the standard amino acids and is symmetric; higher scores reflect more favorable substitutions, with the diagonal containing the largest values. Substitutions that are rare or unlikely have negative scores.

BLOSUM62 is widely used as the default scoring matrix in many protein alignment tools, notably BLASTP, because it provides a balance between sensitivity for detecting distant relationships and specificity for closer ones. It is also used for various sequence similarity searches and alignments in proteomics and molecular biology.

There are other matrices in the BLOSUM family with different identity thresholds, such as BLOSUM45 and BLOSUM80, chosen for different evolutionary distances. While BLOSUM62 is a general-purpose choice, it has limitations: it assumes constant substitution frequencies across contexts and may be less effective for highly divergent sequences or very short alignments. In many modern workflows, profile-based methods (e.g., HMMs, PSI-BLAST) can outperform single matrices.