LnLCorr

LnLCorr detects pairwise coevolutionary relationships among amino acid residues in protein sequences using phylogeny-based maximum likelihood models and likelihood ratio tests.

Key Features:

• Phylogeny-Based Maximum Likelihood Method: Uses phylogeny-aware maximum likelihood to model sequence evolution while accounting for phylogenetic relationships and variation in evolutionary rates across branches.
• Likelihood Ratio Tests: Applies likelihood ratio tests to identify statistically significant coevolution between site pairs.
• Two-State System Reduction: Reduces each site to a two-state representation irrespective of observed residue diversity to simplify the detection of pairwise coevolution.
• Physicochemical Grouping: Categorizes residues by size, charge, and other physicochemical properties to detect interaction patterns including charge-based correlations.
• Context-Dependent Analysis: Evaluates context dependence of substitutions and highlights interactions related to functional regions, for example proton channels in cytochrome c oxidase subunit I.
• Simulation-Based Validation: Incorporates simulations to validate detection of simple correlations and to assess data sufficiency for distinguishing true coevolution from background noise.
• Detection of Negative Coevolution: Detects negative coevolutionary signals such as charge-based interactions within alpha-helices.
• Recognition of Proximal and Distal Interactions: Identifies both spatially proximal and distal coevolving sites on protein surfaces.

Scientific Applications:

• Structural Insights: Identifies coevolving residue pairs that suggest spatial proximity or functional linkage within three-dimensional protein structures.
• Functional Genomics: Illuminates how sequence evolution affects protein function and stability by analyzing coevolutionary patterns across genomic datasets.
• Phylogenetic Reconstruction: Incorporates genomic biodiversity and site-specific interactions to improve inference of evolutionary processes.
• Protein Stability: Reveals long-distance residue interactions important for maintaining protein stability and evolutionary constraints on protein design.
• Adaptive Evolution: Helps distinguish coevolution from adaptive substitutions by parsing context-dependent substitution probabilities.
• Coevolutionary Networks: Facilitates analysis of coevolutionary networks within protein complexes and subunit interactions.

Methodology:

Applies phylogeny-based maximum likelihood models and likelihood ratio tests; reduces sites to two-state representations; groups residues by physicochemical properties; performs context-dependent analyses; and uses simulations to validate correlations and assess data sufficiency.