Many biochemical and biomedical applications such as protein engineering or drug design are concerned with finding functionally similar proteins, however, this remains to be a challenging task. We present a new image-based approach for identifying and visually comparing proteins with similar function that builds on the hierarchical clustering of Molecular Surface Maps. Such maps are two-dimensional representations of complex molecular surfaces and can be used to visualize the topology and different physico-chemical properties of proteins. Our method is based on the idea that visually similar maps also imply a similarity in the function of the mapped proteins. To determine map similarity, we compute descriptive feature vectors using image moments, color moments, or a Convolutional Neural Network and use them for a hierarchical clustering of the maps. We demonstrate the feasibility of our approach using two data sets: an ensemble of hand-selected proteins with known similarities used for verification and an ensemble of ketolase enzymes, where we analyzed the individual domains using our method. Our method is integrated in an interactive visualization application, which allows users to explore and analyze the results. It visualizes the hierarchical clustering and offers linked views that provide details for a comparative data analysis.
%0 Journal Article
%1 schatz2021analyzing
%A Schatz, Karsten
%A Frieß, Florian
%A Schäfer, Marco
%A Buchholz, Patrick C. F.
%A Pleiss, Jürgen
%A Ertl, Thomas
%A Krone, Michael
%D 2021
%J Computers & Graphics
%K myown visus:friessfn from:karstenschatz vis(us) visus:ertl visus visus:schatzkn
%P 114-127
%R 10.1016/j.cag.2021.06.007
%T Analyzing the similarity of protein domains by clustering Molecular Surface Maps
%U https://www.sciencedirect.com/science/article/abs/pii/S0097849321001229
%V 99
%X Many biochemical and biomedical applications such as protein engineering or drug design are concerned with finding functionally similar proteins, however, this remains to be a challenging task. We present a new image-based approach for identifying and visually comparing proteins with similar function that builds on the hierarchical clustering of Molecular Surface Maps. Such maps are two-dimensional representations of complex molecular surfaces and can be used to visualize the topology and different physico-chemical properties of proteins. Our method is based on the idea that visually similar maps also imply a similarity in the function of the mapped proteins. To determine map similarity, we compute descriptive feature vectors using image moments, color moments, or a Convolutional Neural Network and use them for a hierarchical clustering of the maps. We demonstrate the feasibility of our approach using two data sets: an ensemble of hand-selected proteins with known similarities used for verification and an ensemble of ketolase enzymes, where we analyzed the individual domains using our method. Our method is integrated in an interactive visualization application, which allows users to explore and analyze the results. It visualizes the hierarchical clustering and offers linked views that provide details for a comparative data analysis.
@article{schatz2021analyzing,
abstract = {Many biochemical and biomedical applications such as protein engineering or drug design are concerned with finding functionally similar proteins, however, this remains to be a challenging task. We present a new image-based approach for identifying and visually comparing proteins with similar function that builds on the hierarchical clustering of Molecular Surface Maps. Such maps are two-dimensional representations of complex molecular surfaces and can be used to visualize the topology and different physico-chemical properties of proteins. Our method is based on the idea that visually similar maps also imply a similarity in the function of the mapped proteins. To determine map similarity, we compute descriptive feature vectors using image moments, color moments, or a Convolutional Neural Network and use them for a hierarchical clustering of the maps. We demonstrate the feasibility of our approach using two data sets: an ensemble of hand-selected proteins with known similarities used for verification and an ensemble of ketolase enzymes, where we analyzed the individual domains using our method. Our method is integrated in an interactive visualization application, which allows users to explore and analyze the results. It visualizes the hierarchical clustering and offers linked views that provide details for a comparative data analysis.},
added-at = {2021-08-09T19:34:26.000+0200},
author = {Schatz, Karsten and Frieß, Florian and Schäfer, Marco and Buchholz, Patrick C. F. and Pleiss, Jürgen and Ertl, Thomas and Krone, Michael},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/2bf099743f2acfaccfd3fe389843b94ea/visus},
doi = {10.1016/j.cag.2021.06.007},
interhash = {6f88f4a3280ce0042b8ae6dadc5ee994},
intrahash = {bf099743f2acfaccfd3fe389843b94ea},
journal = {Computers & Graphics},
keywords = {myown visus:friessfn from:karstenschatz vis(us) visus:ertl visus visus:schatzkn},
pages = {114-127},
timestamp = {2021-08-09T17:34:26.000+0200},
title = {Analyzing the similarity of protein domains by clustering Molecular Surface Maps},
url = {https://www.sciencedirect.com/science/article/abs/pii/S0097849321001229},
volume = 99,
year = 2021
}