MAVL/StickWRLD: analyzing structural constraints using interpositional dependencies in biomolecular sequence alignments

The increasing availability of structurally aligned protein families has made it possible to use statistical methods to discover regions of interpositional dependencies of residue identity. Such dependencies amongst residues often have structural or functional implications, and their discovery can supply valuable constraints that assist in the refinement of measured, or predicted molecular structure assignments. Multiple Alignment Variation Linker (MAVL) and StickWRLD [W. Ray (2004) Nucleic Acids Res., 32, W59-W63] were developed to analyze and visualize nucleic acid and protein alignments, to discover and illuminate position/location relationships to the user. The original system analyzed users' data from a web-form submission and presented the user with a static VRML diagram describing their data. We are pleased to report that MAVL/StickWRLD has been completely redesigned and rewritten. MAVL/StickWRLD now functions as a platform-independent Java applet, with real-time dynamic controls that enable much more intuitive exploration and interaction with the data. The system has also been upgraded to enable visualization of a range of aggregate residue properties, and an extensive database of pre-computed StickWRLD diagrams based on PFAM families is now available directly from the interface. The Java StickWRLD applet is available via the WWW at http://www.microbial-pathogenesis.org/stickwrld/.