Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

A hierarchical elastic network model for unsupervised electron density map segmentation

Burger, Virginia and Bahar, Ivet and Chennubhotla, Chakra (2011) A hierarchical elastic network model for unsupervised electron density map segmentation. In: Pacific Symposium on Biocomputing (PSB) 2011.

[img]
Preview
PDF
Published Version
Available under License : See the attached license file.

Download (1MB) | Preview
[img] Plain Text (licence)
Available under License : See the attached license file.

Download (1kB)

Abstract

To highlight different levels of structural and functional organization found in cryo EM density maps, we use an unsupervised hierarchical elastic network model (hENM) based on a Markov diffusion process. The hENM segments electron density maps into meaningful protein subregions Chennubhotla & Bahar 2006, and finds a simplified representation of the map in each level of the hierarchy. When implemented on maps in the 2010 Cryo-EM Modeling Challenge, hENM segmented the molecules into individual proteins and regions without manual input. For example, hENM identified the apical and equatorial domains of the 14 groEL proteins in the 23.5A groEL-groES map (EMD: 1046). These results suggest that hENM could be applied to segment density maps which otherwise would require manual curating. As the sub-regions found in each level of the hierarchy highlight the core structural regions of the protein, they can also be used as anchor points for fitting high resolution structures into EM maps. Inference of secondary structure from characteristic patterns in the affinity maps can also help guide the fitting process. Further implementations of our method will study protein dynamics at the coarsest hierarchy level, and then project insight into dynamics onto finer, more detailed levels. This method provides a unifying framework in allowing protein segmentation concurrently with the study of information propagation through the network, with dynamical inference as an additional benefit.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: Conference or Workshop Item (Poster)
Status: Published
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Burger, Virginia
Bahar, Ivetbahar@pitt.eduBAHAR
Chennubhotla, Chakra
Date: 7 February 2011
Date Type: Publication
Journal or Publication Title: F1000 Research
Publisher: F1000 Research Ltd.
Event Title: Pacific Symposium on Biocomputing (PSB) 2011
Event Type: Conference
Schools and Programs: School of Medicine > Computational and Systems Biology
Refereed: No
ISSN: 2046-1402
Official URL: https://f1000research.com/posters/807
Additional Information: Competing Interests: No relevant conflicts of interest declared.
Date Deposited: 22 Nov 2016 17:39
Last Modified: 01 Nov 2017 14:02
URI: http://d-scholarship.pitt.edu/id/eprint/30149

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item