Separation of Sub-Image Entities in Cryo-Electron Microscopy Density Maps by Clustering
Description/Abstract/Artist Statement
Cryo-Electron Microscopy (cryo-EM) is a biophysical technique to produce 3-dimensional electron density maps. Existing methods identify points involved in secondary structures of proteins from a cryo-EM density map at a medium resolution, such as 5-10Å. Due to the close proximity of multiple secondary structures in the density map, computational analysis is needed to identify and separate the secondary structure entities in such maps. We use a modified form of the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm to establish clusters based on spatial and non-spatial proximity. Our current method works well in regions that have clear separation between them. Our future work aims to separate structures in close proximity based on proximity in electron density values, the expected geometrical shape of secondary structures, and the spatial density. By separating the secondary structures into clusters, we can identify sub-regions of the 3-dimensional image that belong to different types of secondary structures, such as α-helices and β-sheets.
Faculty Advisor/Mentor
Jing He
Presentation Type
Oral Presentation
Disciplines
Bioinformatics | Structural Biology | Theory and Algorithms
Session Title
Chemistry
Location
Learning Commons @ Perry Library Conference Room 1310
Start Date
2-2-2019 11:30 AM
End Date
2-2-2019 12:30 PM
Separation of Sub-Image Entities in Cryo-Electron Microscopy Density Maps by Clustering
Learning Commons @ Perry Library Conference Room 1310
Cryo-Electron Microscopy (cryo-EM) is a biophysical technique to produce 3-dimensional electron density maps. Existing methods identify points involved in secondary structures of proteins from a cryo-EM density map at a medium resolution, such as 5-10Å. Due to the close proximity of multiple secondary structures in the density map, computational analysis is needed to identify and separate the secondary structure entities in such maps. We use a modified form of the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm to establish clusters based on spatial and non-spatial proximity. Our current method works well in regions that have clear separation between them. Our future work aims to separate structures in close proximity based on proximity in electron density values, the expected geometrical shape of secondary structures, and the spatial density. By separating the secondary structures into clusters, we can identify sub-regions of the 3-dimensional image that belong to different types of secondary structures, such as α-helices and β-sheets.