Automated function prediction in protein structures

WebFEATURE is a web-accessible structural analysis tool that allows users to scan structures for functional sites in both proteins and nucleic acids. WebFEATURE is the public interface to FEATURE, software that creates computable representations of 3D physicochemical motifs in molecular structures. WebFEATURE 4.0 uses Naive Bayes and Support Vector Machine to train protein functional models and predict active sites in protein structures.

Given an input structure or Protein Data Bank identifier (PDB ID), and a statistical model of a functional site, WebFEATURE will return rank-scored 'hits' in 3D space that identify regions in the structure where similar distributions of physicochemical properties occur relative to the site model. More than a hundred functional site models are available for use, some manually but most automatically built using the SeqFEATURE framework [Wu et al. 2008], where models are seeded from sequence motifs. Users can visualize and interactively manipulate scored hits and the query structure in web browsers that support the Chime plug-in. Alternatively, results can be downloaded and visualized through other freely available molecular modeling tools, like RasMol, PyMOL and Chimera. A major application of FEATURE is in rapid annotation of function to structures in the context of structural genomics, and WebFEATURE may help in this endeavor.

WebFEATURE contains ready-made models and a function scanning interface. This data currently can be downloaded in full or separately by model as flat files. Users wishing to construct their own models need only download the FEATURE package. We are currently re-scanning a 2013 snapshot of the PDB with updated models and will have this data available shortly.

(For more detailed descriptions of methods, see the Methods page, and our publications. To read more about FEATURE and associated projects, see the FEATURE page.)