Persistent topology for cryo-EM data analysis.

摘要: Summary In this work, we introduce persistent homology for the analysis of cryo-electron microscopy (cryo-EM) density maps. We identify topological fingerprint or signature noise, which is widespread in cryo-EM data. For low signal-to-noise ratio (SNR) volumetric data, intrinsic features biomolecular structures are indistinguishable from noise. To remove employ geometric flows that found to preserve fingerprints and diminish In particular, enables us visualize gradual separation those noise during denoising process, gives rise a practical procedure prescribing threshold extract structure information contaminated data after certain iterations flow equation. further demonstrate utility analysis, consider microtubule intermediate Electron Microscopy Data (EMD 1129). Three helix models, an alpha-tubulin monomer model, beta-tubulin dimer constructed fit The least square fitting leads similarly high correlation coefficients, indicates determination via optimization ill-posed inverse problem. However, these models have dramatically different fingerprints. Especially, linkages connectivities discriminate one model another, play little role traditional but very sensitive crucial identified denoising. By comparison original three third topologically favored. present work offers based new strategies resolving problems. Copyright © 2015 John Wiley & Sons, Ltd.