Thumbnails:
List:
Year:
Category:
Session:
Poster:
Getting poster data...
William Ray (Nationwide Children's Hospital, Columbus Ohio, USA; The Ohio State University, Columbus Ohio, USA)
Proteins are 3-dimensional molecules with functions defined by their spatial presentation of physical properties to their environment. The canonical bioinformatic representation by which proteins are studied however, is a sequential listing of their subunits. Since the earliest days of Alanine-scanning mutagenesis, it has been obvious that protein residues do not contribute to function in a purely sequential fashion. As a result, sequence representations map poorly to the actual properties of molecules. This problem is not simply a conceptual one, but a pernicious one as well. The use of a sequential representation engenders the application of analysis methods that are best suited to sequential data, and which capture non-sequential features poorly, if at all. Where such features are not completely disguised, analysis and understanding of non-sequential features requires exponentially more effort when constrained by sequential representations. Herein we describe StickWRLD, a Visual Analytics alternative to purely sequential representations. StickWRLD is an interactive tool that conveys both the sequential and the dependency characteristics of a protein sequence family.