Limits...
Validation of 3D EM Reconstructions: The Phantom in the Noise.

Heymann JB - AIMS Biophys (2015)

Bottom Line: This poses the risk of inappropriate data processing with dubious results.How can we test that a map is a coherent structure present in the images selected from the micrographs?Adopting such a test can aid the microscopist in assessing the usefulness of the micrographs taken before committing to lengthy processing with questionable outcomes.

View Article: PubMed Central - HTML - PubMed

Affiliation: National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, 50 South Dr, Bethesda, MD 20892, USA.

ABSTRACT

Validation is a necessity to trust the structures solved by electron microscopy by single particle techniques. The impressive achievements in single particle reconstruction fuel its expansion beyond a small community of image processing experts. This poses the risk of inappropriate data processing with dubious results. Nowhere is it more clearly illustrated than in the recovery of a reference density map from pure noise aligned to that map-a phantom in the noise. Appropriate use of existing validating methods such as resolution-limited alignment and the processing of independent data sets ("gold standard") avoid this pitfall. However, these methods can be undermined by biases introduced in various subtle ways. How can we test that a map is a coherent structure present in the images selected from the micrographs? In stead of viewing the phantom emerging from noise as a cautionary tale, it should be used as a defining baseline. Any map is always recoverable from noise images, provided a sufficient number of images are aligned and used in reconstruction. However, with smaller numbers of images, the expected coherence in the real particle images should yield better reconstructions than equivalent numbers of noise or background images, even without masking or imposing resolution limits as potential biases. The validation test proposed is therefore a simple alignment of a limited number of micrograph and noise images against the final reconstruction as reference, demonstrating that the micrograph images yield a better reconstruction. I examine synthetic cases to relate the resolution of a reconstruction to the alignment error as a function of the signal-to-noise ratio. I also administered the test to real cases of publicly available data. Adopting such a test can aid the microscopist in assessing the usefulness of the micrographs taken before committing to lengthy processing with questionable outcomes.

No MeSH data available.


Resolution estimates (FSC0.5 values) for reconstructions from microscopic images (blue) for (a) KLH: size 2402 at 2.2 Å/pixel, symmetry D5, and (b) P-SSP7: size 5762 at 1.17 Å/pixel, icosahedral. The published resolutions are shown in red. Also shown are resolution estimates for aligned gaussian noise (gray) and background images (green). Each point is the average of resolution estimates of 10 reconstructions with standard deviations as indicated by the error bars.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4440490&req=5

Figure 4: Resolution estimates (FSC0.5 values) for reconstructions from microscopic images (blue) for (a) KLH: size 2402 at 2.2 Å/pixel, symmetry D5, and (b) P-SSP7: size 5762 at 1.17 Å/pixel, icosahedral. The published resolutions are shown in red. Also shown are resolution estimates for aligned gaussian noise (gray) and background images (green). Each point is the average of resolution estimates of 10 reconstructions with standard deviations as indicated by the error bars.

Mentions: The resolution estimates for reconstructions from particle images are much better than those for noise images from the same number of images (Figure 4). Of note, the behavior of micrograph background and gaussian noise images are very similar, suggesting that the noise images are reasonable substitutes for the background (maybe not surprising [30]). Towards larger numbers, the curves converge in both cases, and then diverge again. This is a reflection of the orientational information in the structure and is highly case-specific. The resolutions for the final KLH and P-SSP7 structures are significantly different from that of the corresponding noise-derived reconstructions. These are examples where the validity of the reconstructions is not in doubt.


Validation of 3D EM Reconstructions: The Phantom in the Noise.

Heymann JB - AIMS Biophys (2015)

Resolution estimates (FSC0.5 values) for reconstructions from microscopic images (blue) for (a) KLH: size 2402 at 2.2 Å/pixel, symmetry D5, and (b) P-SSP7: size 5762 at 1.17 Å/pixel, icosahedral. The published resolutions are shown in red. Also shown are resolution estimates for aligned gaussian noise (gray) and background images (green). Each point is the average of resolution estimates of 10 reconstructions with standard deviations as indicated by the error bars.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4440490&req=5

Figure 4: Resolution estimates (FSC0.5 values) for reconstructions from microscopic images (blue) for (a) KLH: size 2402 at 2.2 Å/pixel, symmetry D5, and (b) P-SSP7: size 5762 at 1.17 Å/pixel, icosahedral. The published resolutions are shown in red. Also shown are resolution estimates for aligned gaussian noise (gray) and background images (green). Each point is the average of resolution estimates of 10 reconstructions with standard deviations as indicated by the error bars.
Mentions: The resolution estimates for reconstructions from particle images are much better than those for noise images from the same number of images (Figure 4). Of note, the behavior of micrograph background and gaussian noise images are very similar, suggesting that the noise images are reasonable substitutes for the background (maybe not surprising [30]). Towards larger numbers, the curves converge in both cases, and then diverge again. This is a reflection of the orientational information in the structure and is highly case-specific. The resolutions for the final KLH and P-SSP7 structures are significantly different from that of the corresponding noise-derived reconstructions. These are examples where the validity of the reconstructions is not in doubt.

Bottom Line: This poses the risk of inappropriate data processing with dubious results.How can we test that a map is a coherent structure present in the images selected from the micrographs?Adopting such a test can aid the microscopist in assessing the usefulness of the micrographs taken before committing to lengthy processing with questionable outcomes.

View Article: PubMed Central - HTML - PubMed

Affiliation: National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, 50 South Dr, Bethesda, MD 20892, USA.

ABSTRACT

Validation is a necessity to trust the structures solved by electron microscopy by single particle techniques. The impressive achievements in single particle reconstruction fuel its expansion beyond a small community of image processing experts. This poses the risk of inappropriate data processing with dubious results. Nowhere is it more clearly illustrated than in the recovery of a reference density map from pure noise aligned to that map-a phantom in the noise. Appropriate use of existing validating methods such as resolution-limited alignment and the processing of independent data sets ("gold standard") avoid this pitfall. However, these methods can be undermined by biases introduced in various subtle ways. How can we test that a map is a coherent structure present in the images selected from the micrographs? In stead of viewing the phantom emerging from noise as a cautionary tale, it should be used as a defining baseline. Any map is always recoverable from noise images, provided a sufficient number of images are aligned and used in reconstruction. However, with smaller numbers of images, the expected coherence in the real particle images should yield better reconstructions than equivalent numbers of noise or background images, even without masking or imposing resolution limits as potential biases. The validation test proposed is therefore a simple alignment of a limited number of micrograph and noise images against the final reconstruction as reference, demonstrating that the micrograph images yield a better reconstruction. I examine synthetic cases to relate the resolution of a reconstruction to the alignment error as a function of the signal-to-noise ratio. I also administered the test to real cases of publicly available data. Adopting such a test can aid the microscopist in assessing the usefulness of the micrographs taken before committing to lengthy processing with questionable outcomes.

No MeSH data available.