fPOP

fPOP catalogs and analyzes protein functional surfaces by geometric comparison of binding-site shapes across holo and apo protein structures for studying protein-ligand interactions.

Key Features:

• Shape Analysis of Binding Sites: Employs a purely geometric method to identify spatial patterns of functional surfaces ("split pockets") in approximately 19,000 bound (holo) structures drawn from a collection of ~42,700 protein structures.
• Prediction and Comparison: Uses identified functional surfaces as spatial templates to predict binding sites on unbound (apo) structures by pairwise alignment and footprinting of pocket fragments against the SplitPocket repository, applying the Smith-Waterman algorithm for shape comparison to assess local geometric similarity.
• Extensive Database: Aggregates roughly 90,000 identified or predicted functional surfaces across approximately 42,700 holo and apo protein structures.

Scientific Applications:

• Functional Surface Study: Enables investigation of structural and functional characteristics of protein binding sites through geometric comparison of pocket templates.
• Conformational Change Assessment: Facilitates evaluation of conformational changes between bound and unbound protein forms by comparing corresponding pocket geometries.
• Functional Divergence Analysis: Supports comparison of functional surfaces across proteins to analyze functional divergence.
• Protein Function Inference: Aids inference of protein function by exploring physicochemical textures and geometric similarity of binding pockets.
• Protein Family Classification: Provides a basis for classifying proteins into families according to shared functional surface geometries.

Methodology:

Geometric identification of split pockets in bound structures followed by pairwise geometric matching and footprinting of unbound pocket fragments against the SplitPocket database using the Smith-Waterman algorithm for shape comparison to assess local structural similarity.