We present a technique that conveys the uncertainty in the secondary structure of proteins-an abstraction model based on atomic coordinates. While protein data inherently contains uncertainty due to the acquisition method or the simulation algorithm, we argue that it is also worth investigating uncertainty induced by analysis algorithms that precede visualization. Our technique helps researchers investigate differences between multiple secondary structure assignment methods. We modify established algorithms for fuzzy classification and introduce a discrepancy-based approach to project an ensemble of sequences to a single importance-weighted sequence. In 2D, we depict the aggregated secondary structure assignments based on the per-residue deviation in a collapsible sequence diagram. In 3D, we extend the ribbon diagram using visual variables such as transparency, wave form, frequency, or amplitude to facilitate qualitative analysis of uncertainty. We evaluated the effectiveness and acceptance of our technique through expert reviews using two example applications: the combined assignment against established algorithms and time-dependent structural changes originating from simulated protein dynamics.
Description
Uncertainty Visualization for Secondary Structures of Proteins
%0 Conference Paper
%1 Schulz2018Uncertainty
%A Schulz, Christoph
%A Schatz, Karsten
%A Krone, Michael
%A Braun, Matthias
%A Ertl, Thomas
%A Weiskopf, Daniel
%B Proceedings of the IEEE Pacific Visualization Symposium (PacificVis)
%D 2018
%I IEEE
%K A01 from:leonkokkoliadis visus:weiskopf sfbtrr161 from:mueller visus:braunms visus:ertl visus visus:schatzkn visus:schulzch
%P 96-105
%R 10.1109/PacificVis.2018.00020
%T Uncertainty Visualization for Secondary Structures of Proteins
%U https://ieeexplore.ieee.org/document/8365980
%X We present a technique that conveys the uncertainty in the secondary structure of proteins-an abstraction model based on atomic coordinates. While protein data inherently contains uncertainty due to the acquisition method or the simulation algorithm, we argue that it is also worth investigating uncertainty induced by analysis algorithms that precede visualization. Our technique helps researchers investigate differences between multiple secondary structure assignment methods. We modify established algorithms for fuzzy classification and introduce a discrepancy-based approach to project an ensemble of sequences to a single importance-weighted sequence. In 2D, we depict the aggregated secondary structure assignments based on the per-residue deviation in a collapsible sequence diagram. In 3D, we extend the ribbon diagram using visual variables such as transparency, wave form, frequency, or amplitude to facilitate qualitative analysis of uncertainty. We evaluated the effectiveness and acceptance of our technique through expert reviews using two example applications: the combined assignment against established algorithms and time-dependent structural changes originating from simulated protein dynamics.
@inproceedings{Schulz2018Uncertainty,
abstract = {We present a technique that conveys the uncertainty in the secondary structure of proteins-an abstraction model based on atomic coordinates. While protein data inherently contains uncertainty due to the acquisition method or the simulation algorithm, we argue that it is also worth investigating uncertainty induced by analysis algorithms that precede visualization. Our technique helps researchers investigate differences between multiple secondary structure assignment methods. We modify established algorithms for fuzzy classification and introduce a discrepancy-based approach to project an ensemble of sequences to a single importance-weighted sequence. In 2D, we depict the aggregated secondary structure assignments based on the per-residue deviation in a collapsible sequence diagram. In 3D, we extend the ribbon diagram using visual variables such as transparency, wave form, frequency, or amplitude to facilitate qualitative analysis of uncertainty. We evaluated the effectiveness and acceptance of our technique through expert reviews using two example applications: the combined assignment against established algorithms and time-dependent structural changes originating from simulated protein dynamics.},
added-at = {2020-10-09T12:31:50.000+0200},
author = {Schulz, Christoph and Schatz, Karsten and Krone, Michael and Braun, Matthias and Ertl, Thomas and Weiskopf, Daniel},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/270436512ee33682c3ee2b91b06a72de7/visus},
booktitle = {Proceedings of the IEEE Pacific Visualization Symposium (PacificVis)},
description = {Uncertainty Visualization for Secondary Structures of Proteins},
doi = {10.1109/PacificVis.2018.00020},
interhash = {a0fda330dcf0712450cfd6fa15eeb884},
intrahash = {70436512ee33682c3ee2b91b06a72de7},
keywords = {A01 from:leonkokkoliadis visus:weiskopf sfbtrr161 from:mueller visus:braunms visus:ertl visus visus:schatzkn visus:schulzch},
pages = {96-105},
publisher = {IEEE},
timestamp = {2020-10-09T10:31:50.000+0200},
title = {Uncertainty Visualization for Secondary Structures of Proteins},
url = {https://ieeexplore.ieee.org/document/8365980},
year = 2018
}