DNA Ribbons

Made for a biology lab at UC Berkeley, using graph solvers to render beautiful traces of the genes that encode for proteins with similar functions. Data is collected from HMMs (Hidden Markov Models) trained to analyze and isolate genes which encode proteins with similar function.

A multi-threaded web-application using web workers to parse data and a proprietary graph solver to create beautiful plots. The plots are rendered using the OffscreenCanvas API from a worker thread and kept responsive using a shared memory layer that can stop expensive tasks mid computation when settings are changed.

Many settings are available to manipulate the plots, highting specific HMM groupings, and a responsive drag-and-drop UI allows organisms to be rearranged. Plots can be exported as high quality pngs ready for academic use, though this tool sees most of its use as a quick way to validate, and rapidly prototype, on modifications to the HMM.