Bulletin of the American Physical Society
2023 APS March Meeting
Volume 68, Number 3
Las Vegas, Nevada (March 5-10)
Virtual (March 20-22); Time Zone: Pacific Time
Session M36: Machine Learning in Scanning Probe Microscopy
8:00 AM–11:00 AM,
Wednesday, March 8, 2023
Room: Room 236
Sponsoring
Unit:
GIMS
Chair: Yulia Maximenko, National Institute of Standards and Tech
Abstract: M36.00001 : TopoStats, a tool to discover the hidden structures and states of biomolecule
8:00 AM–8:36 AM
Presenter:
Alice Pyne
(University of Sheffield, UK)
Author:
Alice Pyne
(University of Sheffield, UK)
However, the lack of automated analysis tools in AFM, and slow integration of machine learning (ML) pipelines limits the analysis of its powerful molecular data. Limiting factors for the design and integration of these tools are AFM specific issues with raw data, and small datasets (compared to e.g. Cryo EM). Raw AFM images must undergo many “cleaning” steps before molecule identification can occur.
We have developed TopoStats (www.github.com/AFM-SPM/TopoStats), an open-source Python utility that handles data cleaning/processing and identifies/characterises individual (bio)molecules, from DNA origami to nuclear pore complexes. This enables us to begin using AFM to generate big data on the structure and conformational state of individual (bio)molecules. We have recently refactored TopoStats to make it easier to use, and to support the majority of AFM file formats.
TopoStats however still currently relies on a pipeline formed of thresholding methods to identify molecules of interest (MoIs). These methods are challenging to generalise and struggle to identify overlapping or more complex structures. We have begun to implement machine learning methodologies, including weakly-supervised random forests, and a recursive DBSCAN algorithm. We demonstrate that we can use these to identify multiple MoIs in one pass with higher accuracy and less user oversight than the gold-standard[2,3] software.
[1] Pyne, A. L. B. et al. Nature Communications 12, 1053 (2021).
[2] Necas, D. & Klapetek, P. Central European Journal of Physics 10, 181–188 (2011)
[3] Beton, J. G. et al. Methods 193, 68–79 (2021).
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2025 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700