Rapid reconstruction of SARS-CoV-2 using a synthetic genomics platform.

Reverse genetics has been an indispensable tool to gain insights into viral pathogenesis and vaccine development. The genomes of large RNA viruses, such as those from coronaviruses, are cumbersome to clone and manipulate in Escherichia coli owing to the size and occasional instability of the genome1-3. Therefore, an alternative rapid and robust reverse-genetics platform for RNA viruses would benefit the research community. Here we show the full functionality of a yeast-based synthetic genomics platform to genetically reconstruct diverse RNA viruses, including members of the Coronaviridae, Flaviviridae and Pneumoviridae families. Viral subgenomic fragments were generated using viral isolates, cloned viral DNA, clinical samples or synthetic DNA, and these fragments were then reassembled in one step in Saccharomyces cerevisiae using transformation-associated recombination cloning to maintain the genome as a yeast artificial chromosome. T7 RNA polymerase was then used to generate infectious RNA to rescue viable virus. Using this platform, we were able to engineer and generate chemically synthesized clones of the virus, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)4, which has caused the recent pandemic of coronavirus disease (COVID-19), in only a week after receipt of the synthetic DNA fragments. The technical advance that we describe here facilitates rapid responses to emerging viruses as it enables the real-time generation and functional characterization of evolving RNA virus variants during an outbreak.

Characterization of intrinsic and effective fitness changes caused by temporarily fixed mutations in the SARS-CoV-2 spike E484 epitope and identification of an epistatic precondition for the evolution of E484A in variant Omicron.

BACKGROUND: Intrinsic fitness costs are likely to have guided the selection of lineage-determining mutations during emergence of variants of SARS-CoV-2. Whereas changes in receptor affinity and antibody neutralization have been thoroughly mapped for individual mutations in spike, their influence on intrinsic replicative fitness remains understudied. METHODS: We analyzed mutations in immunodominant spike epitope E484 that became temporarily fixed over the pandemic. We engineered the resulting immune escape mutations E484K, -A, and -Q in recombinant SARS-CoV-2. We characterized viral replication, entry, and competitive fitness with and without immune serum from humans with defined exposure/vaccination history and hamsters monospecifically infected with the E484K variant. We additionally engineered a virus containing the Omicron signature mutations N501Y and Q498R that were predicted to epistatically enhance receptor binding. RESULTS: Multistep growth kinetics in Vero-, Calu-3, and NCI-H1299 were identical between viruses. Synchronized entry experiments based on cold absorption and temperature shift identified only an insignificant trend toward faster entry of the E484K variant. Competitive passage experiments revealed clear replicative fitness differences. In absence of immune serum, E484A and E484Q, but not E484K, were replaced by wildtype (WT) in competition assays. In presence of immune serum, all three mutants outcompeted WT. Decreased E484A fitness levels were over-compensated for by N501Y and Q498R, identifying a putative Omicron founder background that exceeds the intrinsic and effective fitness of WT and matches that of E484K. Critically, the E484A/Q498R/N501Y mutant and E484K have equal fitness also in presence of pre-Omicron vaccinee serum, whereas the fitness gain by E484K is lost in the presence of serum raised against the E484K variant in hamsters. CONCLUSIONS: The emergence of E484A and E484Q prior to widespread population immunity may have been limited by fitness costs. In populations already exposed to the early immune escape epitope E484K, the Omicron founder background may have provided a basis for alternative immune escape evolution via E484A. Studies of major antigenic epitope changes with and without their epistatic context help reconstruct the sequential adjustments of intrinsic fitness versus neutralization escape during the evolution of major SARS-CoV-2 variants in an increasingly immune human population.

Machine learning techniques to characterize functional traits of plankton from image data.

Plankton imaging systems supported by automated classification and analysis have improved ecologists' ability to observe aquatic ecosystems. Today, we are on the cusp of reliably tracking plankton populations with a suite of lab-based and in situ tools, collecting imaging data at unprecedentedly fine spatial and temporal scales. But these data have potential well beyond examining the abundances of different taxa; the individual images themselves contain a wealth of information on functional traits. Here, we outline traits that could be measured from image data, suggest machine learning and computer vision approaches to extract functional trait information from the images, and discuss promising avenues for novel studies. The approaches we discuss are data agnostic and are broadly applicable to imagery of other aquatic or terrestrial organisms.

Assessing Representation Learning and Clustering Algorithms for Computer-Assisted Image Annotation-Simulating and Benchmarking MorphoCluster.

Image annotation is a time-consuming and costly task. Previously, we published MorphoCluster as a novel image annotation tool to address problems of conventional, classifier-based image annotation approaches: their limited efficiency, training set bias and lack of novelty detection. MorphoCluster uses clustering and similarity search to enable efficient, computer-assisted image annotation. In this work, we provide a deeper analysis of this approach. We simulate the actions of a MorphoCluster user to avoid extensive manual annotation runs. This simulation is used to test supervised, unsupervised and transfer representation learning approaches. Furthermore, shrunken k-means and partially labeled k-means, two new clustering algorithms that are tailored specifically for the MorphoCluster approach, are compared to the previously used HDBSCAN*. We find that labeled training data improve the image representations, that unsupervised learning beats transfer learning and that all three clustering algorithms are viable options, depending on whether completeness, efficiency or runtime is the priority. The simulation results support our earlier finding that MorphoCluster is very efficient and precise. Within the simulation, more than five objects per simulated click are being annotated with 95% precision.

IOL Formula Constants: Strategies for Optimization and Defining Standards for Presenting Data.

PURPOSE: The aim of this study is to present strategies for optimization of lens power (IOLP) formula constants and to show options how to present the results adequately. METHODS: A dataset of N = 1,601 preoperative biometric values, IOLP data and postoperative refraction data was split into a training set and a test set using a random sequence. Based on the training set, we calculated the formula constants for established lens calculation formulae with different methods. Based on the test set, we derived the formula prediction error (PE) as difference of the achieved refraction from the formula predicted refraction. RESULTS: For formulae with 1 constant, it is possible to back-calculate the individual constant for each case using formula inversion. However, this is not possible for formulae with >1 constant. In these cases, more advanced concepts such as non-linear optimization strategies are necessary to derive the formula constants. During cross-validation, measures such as the mean absolute or the root mean squared PE or the ratio of cases within mean absolute PE (MAE) limits could be used as quality measures. CONCLUSIONS: Different constant optimization concepts yield different results. To test the performance of optimized formula constants, a cross-validation strategy is mandatory. We recommend performance curves, where the ratio of cases within absolute PE limits is plotted against the MAE.

Fuzzy Overclustering: Semi-Supervised Classification of Fuzzy Labels with Overclustering and Inverse Cross-Entropy.

Deep learning has been successfully applied to many classification problems including underwater challenges. However, a long-standing issue with deep learning is the need for large and consistently labeled datasets. Although current approaches in semi-supervised learning can decrease the required amount of annotated data by a factor of 10 or even more, this line of research still uses distinct classes. For underwater classification, and uncurated real-world datasets in general, clean class boundaries can often not be given due to a limited information content in the images and transitional stages of the depicted objects. This leads to different experts having different opinions and thus producing fuzzy labels which could also be considered ambiguous or divergent. We propose a novel framework for handling semi-supervised classifications of such fuzzy labels. It is based on the idea of overclustering to detect substructures in these fuzzy labels. We propose a novel loss to improve the overclustering capability of our framework and show the benefit of overclustering for fuzzy labels. We show that our framework is superior to previous state-of-the-art semi-supervised methods when applied to real-world plankton data with fuzzy labels. Moreover, we acquire 5 to 10% more consistent predictions of substructures.

On the Chromatic Dispersion of Hydrophobic and Hydrophilic Intraocular Lenses.

SIGNIFICANCE: There is a high variation of chromatic dispersion with contemporary intraocular lens (IOL) materials. It is well known that chromatic aberration limits the optical performance especially with high-power lenses. Lens manufacturers, however, rarely provide data on the chromatic dispersion of their materials, limiting the comparability of available materials. PURPOSE: This study aimed to analyze chromatic dispersion of hydrophobic and hydrophilic IOLs without prior knowledge of the IOLs' geometries. METHODS: We adapted Bessel's method for measuring focal length by placing the IOL in a wet cell. The chromatic dispersion of several hydrophobic and hydrophilic IOLs was characterized by measuring their focal lengths at multiple wavelengths. From the measured focal lengths, the refractive indices and the Abbe numbers were obtained. We measured four hydrophobic and two hydrophilic IOL models with a nominal power of 21 to 29.5 D. RESULTS: The hydrophobic IOLs had lower Abbe numbers (Abbe numbers <41) than did the hydrophilic IOLs (Abbe numbers >50). Most Abbe numbers were in agreement with the values provided by the IOL manufacturers, and the measurements were independent from IOL power. The repeatability for the Abbe number was better than ±3.5% for all lenses and better than ±2% for lenses between 21 and 26.5 D. The dispersion could be described by a Conrady model (R > 0.997). CONCLUSIONS: The hydrophobic materials showed larger dispersion than did the hydrophilic IOL materials resulting in increased chromatic aberration. The method allowed for an estimation of the IOL's Abbe number without prior knowledge of IOL geometry.

MorphoCluster: Efficient Annotation of Plankton Images by Clustering.

In this work, we present MorphoCluster, a software tool for data-driven, fast, and accurate annotation of large image data sets. While already having surpassed the annotation rate of human experts, volume and complexity of marine data will continue to increase in the coming years. Still, this data requires interpretation. MorphoCluster augments the human ability to discover patterns and perform object classification in large amounts of data by embedding unsupervised clustering in an interactive process. By aggregating similar images into clusters, our novel approach to image annotation increases consistency, multiplies the throughput of an annotator, and allows experts to adapt the granularity of their sorting scheme to the structure in the data. By sorting a set of 1.2 M objects into 280 data-driven classes in 71 h (16 k objects per hour), with 90% of these classes having a precision of 0.889 or higher. This shows that MorphoCluster is at the same time fast, accurate, and consistent; provides a fine-grained and data-driven classification; and enables novelty detection.

Impact of intraocular lens displacement on the fixation axis.

To investigate the impact of intraocular lens (IOL) decentration ≤±1 mm and IOL tilt ≤±10° on the fixation axis and spherical equivalent refraction (SE), 50 pseudo-phakic eyes were simulated using numerical ray-tracing. We computed the position of the object point whose image ends up at the virtual fovea for each scenario and estimated the corresponding change of fixation axis and SE. The eye turned opposite to the direction of IOL decentration or tilt to compensate for the associated prismatic effect (angle <1.2°). Decentration of the aspheric IOL resulted in a hyperopic shift (<0.57 D), and tilt in a myopic shift (<0.77 D).

[Interpretation of the Intraocular Lens Constants for the Haigis Formula]. / Interpretation der Intraokularlinsenkonstanten für die Haigis-Formel.

Background The Haigis formula uses a linear regression with three IOL constants for the prediction of the effective lens position (ELP) of the intraocular lens (IOL), ELP ≈ a0 + a1 ACD + a2 L. It is based on the preoperative anterior chamber depth (ACD) and axial length (L). Material and Methods Differences between IOL constant triplets can be judged based on their statistical measurement uncertainty. To investigate, if the estimation of the average ELP with the help of the average ACD and average L according to 〈ELP〉 ≈ a0 + a1 〈ACD〉 + a2 〈L〉 provides a possible alternative, we have compared both methods. The results based on two different strategies for optimisation of the IOL constants a0, a1, a2 are used for illustration. Results The estimation of the average ELP is suitable for basic categorisation of the IOL constants. The confidence-volumes in shape of ellipsoids based on the statistical measurement uncertainties of the IOL constant optimisations allow a better comparison between IOL constant triplets a0, a1, a2.

Protein turnover quantification in a multilabeling approach: from data calculation to evaluation.

Liquid chromatography coupled to tandem mass spectrometry in combination with stable-isotope labeling is an established and widely spread method to measure gene expression on the protein level. However, it is often not considered that two opposing processes are responsible for the amount of a protein in a cell--the synthesis as well as the degradation. With this work, we provide an integrative, high-throughput method--from the experimental setup to the bioinformatics analysis--to measure synthesis and degradation rates of an organism's proteome. Applicability of the approach is demonstrated with an investigation of heat shock response, a well-understood regulatory mechanism in bacteria, on the biotechnologically relevant Corynebacterium glutamicum. Utilizing a multilabeling approach using both heavy stable nitrogen as well as carbon isotopes cells are metabolically labeled in a pulse-chase experiment to trace the labels' incorporation in newly synthesized proteins and its loss during protein degradation. Our work aims not only at the calculation of protein turnover rates but also at their statistical evaluation, including variance and hierarchical cluster analysis using the rich internet application QuPE.

Ratio of torus and equivalent power to refractive cylinder and spherical equivalent in phakic lenses - a Monte-Carlo simulation study.

BACKGROUND: Spherical and astigmatic powers for phakic intraocular lenses are frequently calculated using fixed ratios of phakic lens refractive power to refractive spherical equivalent, and of phakic lens astigmatism to refractive cylinder. In this study, a Monte-Carlo simulation based on biometric data was used to investigate how variations in biometrics affect these ratios, in order to improve the calculation of implantable lens parameters. METHODS: A data set of over sixteen thousand biometric measurements including axial length, phakic anterior chamber depth, and corneal equivalent and astigmatic power was used to construct a multidimensional probability density distribution. From this, we determined the axial position of the implanted lens and estimated the refractive spherical equivalent and refractive cylinder. A generic data model resampled the density distributions and interactions between variables, and the implantable lens power was determined using vergence propagation. RESULTS: 50 000 artificial data sets were used to calculate the phakic lens spherical equivalent and astigmatism required for emmetropization, and to determine the corresponding ratios for these two values. The spherical ratio ranged from 1.0640 to 1.3723 and the astigmatic ratio from 1.0501 to 1.4340. Both ratios are unaffected by the corneal spherical / astigmatic powers, or the refractive cylinder, but show strong correlation with the refractive spherical equivalent, mild correlation with the lens axial position, and moderate negative correlation with axial length. As a simplification, these ratios could be modelled using a bi-variable linear regression based on the first two of these factors. CONCLUSION: Fixed spherical and astigmatic ratios should not be used when selecting high refractive power phakic IOLs as their variation can result in refractive errors of up to ±0.3 D for a 8 D lens. Both ratios can be estimated with clinically acceptable precision using a linear regression based on the refractive spherical equivalent and the axial position.

Corneal back surface power - interpreting keratometer readings and what predictions can tell us.

The classical Javal's rule allows estimation of refractive cylinder from keratometric astigmatism using scaling for vergence transformation, with an additional half dioptre of cylinder against-the-rule. With increasing popularity of toric intraocular lenses it has been shown that keratometric astigmatism does not fully reflect the entire astigmatism of the phakic or pseudophakic eye. Researchers mostly argue that this mismatch is primarily due to astigmatism of the corneal back surface, and some papers propose correction strategies to consider this mismatch with the keratometric values. In this Technical Note we address this issue using a vector analysis and show the consequences of this correction on the front and back surface as well as total astigmatism of the cornea. As examples we focus on the correction strategies proposed by Abulafia and by Savini, frequently used in clinical practice. The main conclusion is that, since corneal tomographers do not systematically show zero total astigmatism in situations where keratometry measures astigmatism against-the-rule of around 3 dioptres, there may be reasons other than the corneal back surface for this mismatch between keratometry and total astigmatism. A number of possible sources of this mismatch are proposed.

Optimal Dataset Sizes for Constant Optimization in Published Theoretical Optical Formulae.

Purpose: To determine the optimal number of data points required for optimization of formulae for classical lens power calculation.Methods: A large dataset of preoperative biometric values was used to assess the convergence of formula constants in a number of established intraocular lens power calculation formulae.Results: In formulae with a single constant, 80-100 clinical data points are sufficient to obtain convergence. The Haigis formula (three constants) requires 200-300 data points although refractive error converges more rapidly.Conclusions: In all formulae, 80-100 clinical data points are sufficient to achieve a stable mean refractive error.

Comparison of corneal elevation and pachymetry measurements made by two state of the art corneal tomographers with different measurement principles.

PURPOSE: To compare corneal tomography measurements (elevation and pachymetry) as made by two corneal tomographers: Pentacam AXL and CASIA 2. MATERIAL AND METHODS: The devices were used in a standard measuring mode. 77 normal eyes were measured five times with both devices. The data maps for anterior and posterior corneal elevation and pachymetry were exported and analyzed. Repeatability and average values were calculated for each valid data point on the exported data maps. We also calculated a corrected repeatability of the elevation data maps by removing rotation, tilt, and decentration through realignment of the elevation measurement of each eye prior to analyzing the variations in the measurement usingthe same method as for the repeatability. RESULTS: Pentacam AXL offered the better (corrected) repeatability for anterior corneal elevation measurements. CASIA 2 offered better repeatability for the pachymetry measurements. The tomographers could not be used interchangeably. The central corneal thickness was measured 9 µm ± 3 µm larger when measured with Pentacam AXL compared to CASIA 2.

Keratoconic eyes with stable corneal tomography could benefit more from custom intraocular lens design than normal eyes.

We investigated whether eyes with keratoconic corneal tomography pattern could benefit more from aberration correction with custom intraocular lenses (IOLs) than normal cataractous eyes despite the effect of misalignment on the correction of aberrations. Custom IOLs (cIOLs) were calculated for twelve normal and twelve keratoconic eyes using personalized numerical ray tracing models. The Stiles-Crawford weighted root-mean-square spot-size (wRMS) at the virtual fovea was evaluated for cIOLs and aberration-neutral IOLs (nIOLs) in a simulated clinical study with 500 virtual IOL implantations per eye and per IOL. IOL misalignment (decentration, tilt, rotation) and pupillary ectopia (4.5 mm iris aperture) were varied upon each virtual implantation. The nIOLs achieved average wRMS of 16.4 ± 4.3 µm for normal, and 92.7 ± 34.4 µm for keratoconic eyes (mean ± standard deviation). The cIOLs reduced the average wRMS to 10.3 ± 5.8 µm for normal, and 28.5 ± 18.6 µm for keratoconic eyes. The cIOLs produced smaller wRMS than nIOLs in most virtual implantations (86.7% for normal and 99.4% for keratoconic eyes). IOL misalignment resulted in larger wRMS variations in the keratoconus group than in the normal group. Custom freeform IOL-optics-design may become a promising option for the correction of advanced aberrations in eyes with non-progressive keratoconic corneal tomography pattern.

Individually Customized IOL Versus Standard Spherical Aberration-Correcting IOL.

PURPOSE: To compare the visual performance of an individually customized intraocular lens (IOL) versus a standard spherical aberration-correcting IOL. METHODS: In this prospective comparative study, 74 eyes of 60 patients scheduled for cataract surgery were randomized in a 2:1 ratio to receive either an individually customized IOL (; HumanOptics AG, Erlangen, Germany; customized group) or an aspheric IOL with a standard correction of spherical aberration (SA) (Tecnis ZCB00; Johnson & Johnson Vision Surgical, Inc., Santa Ana, CA; standardized group). In the customized group, IOL calculation was based on a minimum of a merit function that contained terms representing residual refraction, residual SA, and modulation transfer function. In the standardized group, the IOL was calculated with a routine procedure using the Holladay formula and had a standard SA correction of -0.27 µm. Refraction, visual acuity (far, intermediate, near), photopic and mesopic contrast sensitivity, defocus curve, corneal and ocular spherical aberration, and pupil size were measured 4 weeks and 3 months postoperatively. RESULTS: The customized group comprised 48 eyes of 37 patients and the standardized group 26 eyes of 23 patients. At 3 months, mean total ocular SA (5 mm) was 0.04 ± 0.06 µm in the customized group and -0.01 ± 0.05 µm in the standardized group. Uncorrected distance visual acuity and distance-corrected near visual acuity were statistically significantly better in the customized group. Contrast sensitivity testing yielded significantly better results in the customized group under photopic and mesopic conditions for almost all spatial frequencies. Compared to the standardized group, the defocus curve of the customized group showed a wider plateau surrounding the distance focal point. CONCLUSIONS: With the implantation of an individually optimized aspheric IOL visual performance, especially contrast sensitivity, can be significantly improved compared to a standard aberration-correcting IOL. [J Refract Surg. 2019;35(9):565-574.].

Relationship between effective lens position and axial position of a thick intraocular lens.

PURPOSE: To discuss the impact of intraocular lens-(IOL)-power, IOL-thickness, IOL-shape, corneal power and effective lens position (ELP) on the distance between the anterior IOL vertex (ALP) of a thick IOL and the ELP of its thin lens equivalent. METHODS: We calculated the ALP of a thick IOL in a model eye, which results in the same focal plane as a thin IOL placed at the ELP using paraxial approximation. The model eye included IOL-power (P), ELP, IOL-thickness (Th), IOL-shape-factor (X), and corneal power (DC). The initial values were P = 10 D (diopter: 1 D = 1 m-1), 20 D, 30 D, Th = 0.9 mm, ELP = 5 mm, X = 0, DC = 43 D. The difference between ALP and the ELP was illustrated as a function of each of the model parameters. RESULTS: The ALP of a thick lens has to be placed in front of the ELP for P>0 IOLs to achieve the same optical effect as the thin lens equivalent. The difference ALP-ELP for the initial values is -0.57 mm. Minus power IOLs (ALP-ELP = -0.07 mm, for IOL-power = -5 D) and convex-concave IOLs (ALP-ELP = -0.16 mm, for X = 1) have to be placed further posterior. The corneal power and ELP have less influence, but corneal power cannot be neglected. CONCLUSION: The distance between ELP and ALP primarily depends on IOL-power, IOL-thickness, and shape-factor.

Reproducibility and normal values of static pupil diameters.

PURPOSE: To provide additional information on normal values of static pupil diameter measurements for binocular infrared pupillometry with PupilX, a commercial pupillometer, and assess the reproducibility of this device's measurements. METHODS: The pupil diameters from 91 study participants with normal eyes with an average age of 39.7 years (SD 16.4 years) were measured with PupilX under scotopic (0 lx), mesopic (1 lx), and photopic (16 lx) illumination. To assess the repeatability of the device, each measurement was repeated 5 times. RESULTS: The mean pupil diameters were 6.5 mm (SD 1.3 mm), 5.5 mm (SD 1.2 mm), and 4.03 mm (SD 0.9 mm) under scotopic, mesopic, and photopic illumination. Left and right eyes showed no difference in mean pupil diameters. The mean unsigned anisocoria was 0.26 mm (SD 0.32 mm) under scotopic, 0.26 mm (SD 0.27 mm) under mesopic, and 0.19 mm (SD 0.19 mm) under photopic illumination. The decrease in pupil diameter with age was largest for scotopic (≈0.057 mm/y) and smallest for photopic illumination (≈0.025 mm/y). The repeatability of the pupillometer was better than 0.2 mm. CONCLUSIONS: This study provides reference values for age- and light-related pupil diameters measured with the PupilX digital pupillometer in normal subjects.

Comparison of Corneal Tomography: Repeatability, Precision, Misalignment, Mean Elevation, and Mean Pachymetry.

PURPOSE: To compare corneal tomography and its statistical uncertainty for measurements obtained by three clinically used corneal tomographers: A Scheimpflug camera (Pentacam HR), a swept source optical coherence tomography system (CASIA SS-1000), and Placido ring imaging (TMS-5). MATERIAL AND METHODS: Repeated measurements with all three devices on 34 normal eyes were used to estimate the repeatability, precision, and mean values of corneal elevation and pachymetry within 8 mm diameter. The repeatability (standard deviation) was calculated for each data point of the corneal elevation data-maps of anterior and posterior cornea as well as for the pachymetry data-maps. Uncertainty on the position of the eye at each measurement might contribute to the differences between elevation data-maps. To take this into account, we defined the precision as the standard deviation for the elevation data-maps of anterior and posterior cornea after correction of misalignment-effects (rotation, translation). The mean elevation and pachymetry data-maps were fitted with Zernike polynomials for interdevice-comparison. RESULTS: Pentacam HR offered the best repeatability and precision for the anterior corneal elevation (<3 and <1.6 µm, respectively). CASIA SS-1000 offered good repeatability and precision with high resolution for posterior corneal elevation, and the best repeatability for pachymetry (<3 µm). TMS-5 measured anterior elevation with similar repeatability to CASIA SS-1000 (<6 µm). The data-maps of the three tomographers could not be used interchangeably. The largest differences were observed for pachymetry and posterior corneal elevation data-maps. CONCLUSIONS: Misalignment limited the repeatability of TMS-5 and Pentacam HR, but had little influence on the repeatability of CASIA SS-1000.