X3D protein ribbon models for the visual analysis of co-evolving amino acids structural relationship
The Challenge
The challenge for statistical modeling of protein data sequences (e.g. for pharmaceutical protein engineering) is understanding how the statistical 2D data relates to protein structures. For example, is there a relationship between pairs of amino acids that might be close in 3D space but distant in a protein sequence. Existing protein viewers do not provide an easy to use method for adding visual representation of data relationships. The ProteinX3D project was developed to map statistical relationships found in a protein sequence onto 3D models of the protein.
The Solution
ProteinX3D is a Java program that uses XML data from the worldwide Protein Data Bank (PDB-XML) and the XPATH query language to create XML X3D ribbon model visualizations. The Java program was less than a 1000 lines of user written code. XML and Xpath reduced the amount of internal source code that was needed for data and organization. From a coding perspective, it was straightforward to add graphic elements representing statistical relationships to the X3D XML structures using XPATH to query amino acids coordinates in PDB-XML data.
The resulting visulizations can be used to detect co-evolving pairs that are close in 3D space but distant in a protein sequence and help predict tertiary structure for that protein family. This in turn allows for the initial relative placement of secondary structures in a tertiary structure and narrows the possible number of protein folding solutions as an accurate base model.
For future analysis, because the structures are modeled in X3D, adding animation or rigid body physics will also be possible without developing a custom application.