FuncPatch

FuncPatch infers conserved functional patches in protein tertiary structures using fast approximate Bayesian phylogenetic methods that account for spatial correlations in substitution rates.

Key Features:

• Approximate Bayesian Inference: Employs an approximate Bayesian framework to infer conserved functional patches as a rapid alternative to computationally intensive methods such as GP4Rate.
• Spatial Correlation Modeling: Integrates spatial correlations in substitution rates across protein tertiary structures, in contrast to methods like Rate4Site that ignore spatial dependence.
• Efficiency and Speed: Achieves orders of magnitude faster runtimes than GP4Rate while maintaining accuracy comparable to more time-consuming Bayesian approaches.
• Complementary Performance: Complements Rate4Site and outperforms 3D sliding window approaches by inferring spatial correlation strength from the data rather than requiring predefined window sizes.

Scientific Applications:

• Structural biology: Identification of conserved functional regions within protein tertiary structures to inform hypotheses about protein function.
• Evolutionary analysis: Detection of spatially correlated substitution-rate patterns for studies of protein evolution and comparative sequence analysis.

Methodology:

Performs fast approximate Bayesian inference within a phylogenetic framework that models spatial correlations in substitution rates across protein tertiary structures; validation was performed using simulations and four empirical case studies corroborated by experimental evidence.