PMB matrix

PMB matrix implements an evolutionary substitution model that derives amino acid substitution probabilities from the Blocks database to improve protein sequence alignment and evolutionary analysis relative to PAM-type and BLOSUM matrices.

Key Features:

• Evolutionary Model Integration: Derived from the Blocks database and incorporating the additivity of evolutionary distances to approximate amino acid substitution probabilities as a function of evolutionary distance.
• Compatibility with BLOSUM Matrices: Maintains compatibility with the BLOSUM series of substitution matrices for use in BLOSUM-based analyses.
• Enhanced Accuracy in Sequence Alignment: Corrects for multiple hits to address limitations of PAM-type and traditional BLOSUM matrices, improving alignment accuracy across varying sequence divergences.
• Implementation in Analysis Software: Implemented within Pseq-gen and usable alongside Ziheng Yang's PAML software for incorporation into computational workflows.

Scientific Applications:

• Phylogenetic analysis: Provides substitution probabilities for phylogenetic inference and model-based sequence evolution studies.
• Comparative genomics: Improves detection of homologs and comparative analyses by using substitution models that reflect true evolutionary distances.
• Functional annotation: Aids inference of conserved residues and functional divergence by modeling amino acid substitutions over evolutionary time scales.

Methodology:

Constructs probabilistic substitution matrices from empirical alignments in the Blocks database, incorporates additivity of evolutionary distances, and corrects for multiple hits to estimate amino acid substitution probabilities.