Accuracy, transferability, and computational efficiency of interatomic potentials for simulations of carbon under extreme conditions.

Large-scale atomistic molecular dynamics (MD) simulations provide an exceptional opportunity to advance the fundamental understanding of carbon under extreme conditions of high pressures and temperatures. However, the fidelity of these simulations depends heavily on the accuracy of classical interatomic potentials governing the dynamics of many-atom systems. This study critically assesses several popular empirical potentials for carbon, as well as machine learning interatomic potentials (MLIPs), in their ability to simulate a range of physical properties at high pressures and temperatures, including the diamond equation of state, its melting line, shock Hugoniot, uniaxial compressions, and the structure of liquid carbon. Empirical potentials fail to accurately predict the behavior of carbon under high pressure-temperature conditions. In contrast, MLIPs demonstrate quantum accuracy, with Spectral Neighbor Analysis Potential (SNAP) and atomic cluster expansion (ACE) being the most accurate in reproducing the density functional theory results. ACE displays remarkable transferability despite not being specifically trained for extreme conditions. Furthermore, ACE and SNAP exhibit superior computational performance on graphics processing unit-based systems in billion atom MD simulations, with SNAP emerging as the fastest. In addition to offering practical guidance in selecting an interatomic potential with a fine balance of accuracy, transferability, and computational efficiency, this work also highlights transformative opportunities for groundbreaking scientific discoveries facilitated by quantum-accurate MD simulations with MLIPs on emerging exascale supercomputers.

Investigating the behavior of diffusion models for accelerating electronic structure calculations.

We present an investigation of diffusion models for molecular generation, with the aim of better understanding how their predictions compare to the results of physics-based calculations. The investigation into these models is driven by their potential to significantly accelerate electronic structure calculations using machine learning, without requiring expensive first-principles datasets for training interatomic potentials. We find that the inference process of a popular diffusion model for de novo molecular generation is divided into an exploration phase, where the model chooses the atomic species, and a relaxation phase, where it adjusts the atomic coordinates to find a low-energy geometry. As training proceeds, we show that the model initially learns about the first-order structure of the potential energy surface, and then later learns about higher-order structure. We also find that the relaxation phase of the diffusion model can be re-purposed to sample the Boltzmann distribution over conformations and to carry out structure relaxations. For structure relaxations, the model finds geometries with ∼10× lower energy than those produced by a classical force field for small organic molecules. Initializing a density functional theory (DFT) relaxation at the diffusion-produced structures yields a >2× speedup to the DFT relaxation when compared to initializing at structures relaxed with a classical force field.

The genetic landscape of antibiotic sensitivity in Staphylococcus aureus.

Despite the critical importance of essential genes, systems-level investigations of their contribution to antibiotic sensitivity have been limited. Using CRISPR Adaptation-mediated Library Manufacturing (CALM), we generated ultra-dense CRISPR interference (CRISPRi) libraries in methicillin-sensitive and -resistant strains of Staphylococcus aureus, which allowed us to quantify gene fitness on a global scale across ten clinically relevant antibiotics. This led to the identification of a comprehensive set of known and novel biological processes modulating bacterial fitness in the antibiotics. Notably, we found that essential genes from diverse processes dominated antibiotic-gene interactions, including a large number of synergistic interactions between bactericidal antibiotics and processes such as cell wall synthesis/cell division (CC), DNA replication/DNA recombination (DD), protein export, and coenzyme A biosynthesis. Simultaneous genetic perturbations of diverse CC and DD processes aggravated bacterial fitness, revealing a widespread synergy between the two highly coordinated processes. In contrast, perturbation of transcriptional, translational, and select energy processes antagonized the effects of bactericidal antibiotics. Finally, we show that small molecule inhibitors recapitulated synergistic antibiotic-gene interactions, providing a rational foundation for developing novel combinatorial antimicrobial therapies.

Cardiovascular Risk Factors and Echocardiographic Findings in a Predominantly Black Population with Rheumatoid Arthritis and Heart Failure.

Among White rheumatoid arthritis (RA) cohorts, heart failure with preserved ejection fraction (HFpEF) is the most prevalent type of heart failure (HF). We aimed to assess the type of HF affecting Black RA patients. 64 patients with RA-HF were compared to age-, sex-, and race-matched RA patients without HF. Left ventricular ejection fraction (LVEF), wall motion abnormalities, left ventricle (LV) mass, and wall thickness were reviewed. 87.3% were Black, 84.4% were women, with a mean age of 69.6 ± 1.38 (± SEM) and BMI (kg/m 2) 29.6 ± 1.07. RA-HF patients had higher rates of hypertension (HTN), chronic kidney disease, and atrial fibrillation. 66.7% had ≥3 cardiovascular risk factors compared to RA patients without HF. 2D-echocardiograms of RA-HF revealed that 62.3% had LVEF ≥50%, 37% had diastolic dysfunction, and 43.1% had wall motion abnormalities. LV mass and relative wall thickness measurements indicated LV eccentric remodeling. The odds ratio for HF was 4.7 (1.5-14.53 CI), p<0.01, among RA-HTN group and 3.5 (1.091-11.7 CI) p<0.01 among smokers. In our predominantly Black RA-HF patients, HFpEF was the most common type of HF. HTN was associated with the highest OR for HF. Eccentric hypertrophic remodeling, a known poor prognostic indicator for cardiovascular events, was found. Further studies are required to confirm our findings.

Extreme Metastability of Diamond and its Transformation to the BC8 Post-Diamond Phase of Carbon.

Diamond possesses exceptional physical properties due to its remarkably strong carbon-carbon bonding, leading to significant resilience to structural transformations at very high pressures and temperatures. Despite several experimental attempts, synthesis and recovery of the theoretically predicted post-diamond BC8 phase remains elusive. Through quantum-accurate multimillion atom molecular dynamics (MD) simulations, we have uncovered the extreme metastability of diamond at very high pressures, significantly exceeding its range of thermodynamic stability. We predict the post-diamond BC8 phase to be experimentally accessible only within a narrow high pressure-temperature region of the carbon phase diagram. The diamond to BC8 transformation proceeds through premelting followed by BC8 nucleation and growth in the metastable carbon liquid. We propose a double-shock compression pathway for BC8 synthesis, which is currently being explored in experiments at the National Ignition Facility.

CathAI: fully automated coronary angiography interpretation and stenosis estimation.

Coronary angiography is the primary procedure for diagnosis and management decisions in coronary artery disease (CAD), but ad-hoc visual assessment of angiograms has high variability. Here we report a fully automated approach to interpret angiographic coronary artery stenosis from standard coronary angiograms. Using 13,843 angiographic studies from 11,972 adult patients at University of California, San Francisco (UCSF), between April 1, 2008 and December 31, 2019, we train neural networks to accomplish four sequential necessary tasks for automatic coronary artery stenosis localization and estimation. Algorithms are internally validated against criterion-standard labels for each task in hold-out test datasets. Algorithms are then externally validated in real-world angiograms from the University of Ottawa Heart Institute (UOHI) and also retrained using quantitative coronary angiography (QCA) data from the Montreal Heart Institute (MHI) core lab. The CathAI system achieves state-of-the-art performance across all tasks on unselected, real-world angiograms. Positive predictive value, sensitivity and F1 score are all ≥90% to identify projection angle and ≥93% for left/right coronary artery angiogram detection. To predict obstructive CAD stenosis (≥70%), CathAI exhibits an AUC of 0.862 (95% CI: 0.843-0.880). In UOHI external validation, CathAI achieves AUC 0.869 (95% CI: 0.830-0.907) to predict obstructive CAD. In the MHI QCA dataset, CathAI achieves an AUC of 0.775 (95%. CI: 0.594-0.955) after retraining. In conclusion, multiple purpose-built neural networks can function in sequence to accomplish automated analysis of real-world angiograms, which could increase standardization and reproducibility in angiographic coronary stenosis assessment.

A Rare Case of a Panic Attack Inducing a Junctional Escape Rhythm on ECG.

We present a unique and rare case of a young female patient who presented with complaints typical of her prior panic attacks and was found to have a junctional escape rhythm on ECG. Upon resolution of her symptoms, a repeat ECG demonstrated a return to normal sinus rhythm. Given that alternative etiologies had been ruled out, it was postulated that her panic attack induced a transient junctional escape rhythm.

Rett-like Phenotypes in HNRNPH2-Related Neurodevelopmental Disorder.

Rett Syndrome (RTT) is a neurodevelopmental disorder with a prevalence of 1:10,000 to 15,000 females worldwide. Classic Rett Syndrome presents in early childhood with a period of developmental regression, loss of purposeful hand skills along with hand stereotypies, gait abnormalities, and loss of acquired speech. Atypical RTT is diagnosed when a child shows some but not all the phenotypes of classic RTT, along with additional supporting criteria. Over 95% of classic RTT cases are attributed to pathogenic variants in Methyl-CpG Binding Protein 2 (MECP2), though additional genes have been implicated in other RTT cases, particularly those with the atypical RTT clinical picture. Other genetic etiologies have emerged with similar clinical characteristics to RTT Syndrome. Our team has characterized HNRNPH2-related neurodevelopmental disorder (HNRNPH2-RNDD) in 33 individuals associated with de novo pathogenic missense variants in the X-linked HNRNPH2 gene, characterized by developmental delay, intellectual disability, seizures, autistic-like features, and motor abnormalities. We sought to further characterize RTT clinical features in this group of individuals by using caregiver report. Twenty-six caregivers completed electronic surveys, with only 3 individuals having previously received an atypical RTT diagnosis, and no individuals with a typical RTT diagnosis. Caregivers reported a high number of behaviors and/or phenotypes consistent with RTT, including the major criteria of the syndrome, such as regression of developmental skills and abnormal gait. Based on the survey results, 12 individuals could meet the diagnostic clinical criteria for atypical RTT Syndrome. In summary, individuals with HNRNPH2-RNDD exhibit clinical characteristics that overlap with those of RTT, and therefore, HNRNPH2-RNDD, should be considered on the differential diagnosis list with this clinical picture.

Spheromers reveal robust T cell responses to the Pfizer/BioNTech vaccine and attenuated peripheral CD8+ T cell responses post SARS-CoV-2 infection.

T cells are a critical component of the response to SARS-CoV-2, but their kinetics after infection and vaccination are insufficiently understood. Using "spheromer" peptide-MHC multimer reagents, we analyzed healthy subjects receiving two doses of the Pfizer/BioNTech BNT162b2 vaccine. Vaccination resulted in robust spike-specific T cell responses for the dominant CD4+ (HLA-DRB1∗15:01/S191) and CD8+ (HLA-A∗02/S691) T cell epitopes. Antigen-specific CD4+ and CD8+ T cell responses were asynchronous, with the peak CD4+ T cell responses occurring 1 week post the second vaccination (boost), whereas CD8+ T cells peaked 2 weeks later. These peripheral T cell responses were elevated compared with COVID-19 patients. We also found that previous SARS-CoV-2 infection resulted in decreased CD8+ T cell activation and expansion, suggesting that previous infection can influence the T cell response to vaccination.

Heterogeneity in IgG-CD16 signaling in infectious disease outcomes.

In this review, we discuss how IgG antibodies can modulate inflammatory signaling during viral infections with a focus on CD16a-mediated functions. We describe the structural heterogeneity of IgG antibody ligands, including subclass and glycosylation that impact binding by and downstream activity of CD16a, as well as the heterogeneity of CD16a itself, including allele and expression density. While inflammation is a mechanism required for immune homeostasis and resolution of acute infections, we focus here on two infectious diseases that are driven by pathogenic inflammatory responses during infection. Specifically, we review and discuss the evolving body of literature showing that afucosylated IgG immune complex signaling through CD16a contributes to the overwhelming inflammatory response that is central to the pathogenesis of severe forms of dengue disease and coronavirus disease 2019 (COVID-19).

Harnessing IgG Fc glycosylation for clinical benefit.

The effector activity of IgG antibodies is regulated at several levels, including IgG subclass, modifications of the Fc glycan, and the distribution of Type I and II Fcγ receptors (FcγR) on effector cells. Here, we explore how Fc glycosylation, particularly sialylation and fucosylation, tunes cellular responses to immune complexes. We review the current understanding of the pathways and mechanisms underlying this biology, address FcγR in antigen presentation, and discuss aspects of the clinical understanding of Fc glycans in therapies and disease.

Antibiotic adverse effects in pediatric acute rhinosinusitis: Systematic review and meta-analysis.

BACKGROUND: Pediatric acute bacterial rhinosinusitis (ABRS) is often treated with oral antibiotics, with limited insight into adverse effects (AEs) across drug classes. In this systematic review and meta-analysis, we characterize AE incidence associated with oral antibiotics in these patients. METHODOLOGY/PRINCIPAL: We searched PubMed and Embase for English-language articles published from 1985 to September 2020 reporting AEs of oral antibiotic therapy for ABRS patients aged 0-18 years. Six-hundred and sixty-six articles underwent title and abstract screening, identifying 154 articles for full-length review. RESULTS: Eleven articles were included, most of which reported individual and aggregate AE incidences. Amoxicillin/clavulanate, amoxicillin, cephalosporin/carbacephem, and placebo groups were identified. Random-effects meta-analysis of prospective groups identified appreciable incidences of diarrhea and abdominal pain, and low incidence of rash, for amoxicillin-clavulanate and amoxicillin. All antibiotics as well as placebo were associated with non-zero overall AE incidence. Children receiving antibiotics were about twice as likely to incur any AE during treatment in placebo-controlled studies, though this association was not significant. High heterogeneity limited most point estimates, with risk of bias, typically in outcomes measurement, detected in most studies. CONCLUSIONS: Reporting of AEs associated with oral antibiotic use in pediatric ABRS is limited in current literature. Adverse effects are non-negligible, but may not significantly exceed placebo.

Early non-neutralizing, afucosylated antibody responses are associated with COVID-19 severity.

A damaging inflammatory response is implicated in the pathogenesis of severe coronavirus disease 2019 (COVID-19), but mechanisms contributing to this response are unclear. In two prospective cohorts, early non-neutralizing, afucosylated immunoglobulin G (IgG) antibodies specific to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were associated with progression from mild to more severe COVID-19. To study the biology of afucosylated IgG immune complexes, we developed an in vivo model that revealed that human IgG-Fc-gamma receptor (FcγR) interactions could regulate inflammation in the lung. Afucosylated IgG immune complexes isolated from patients with COVID-19 induced inflammatory cytokine production and robust infiltration of the lung by immune cells. In contrast to the antibody structures that were associated with disease progression, antibodies that were elicited by messenger RNA SARS-CoV-2 vaccines were highly fucosylated and enriched in sialylation, both modifications that reduce the inflammatory potential of IgG. Vaccine-elicited IgG did not promote an inflammatory lung response. These results show that human IgG-FcγR interactions regulate inflammation in the lung and define distinct lung activities mediated by the IgG that are associated with protection against, or progression to, severe COVID-19.

Antibodies elicited by SARS-CoV-2 infection or mRNA vaccines have reduced neutralizing activity against Beta and Omicron pseudoviruses.

Multiple severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants that have mutations associated with increased transmission and antibody escape have arisen over the course of the current pandemic. Although the current vaccines have largely been effective against past variants, the number of mutations found on the Omicron (B.1.1.529) spike protein appear to diminish the protection conferred by preexisting immunity. Using vesicular stomatitis virus (VSV) pseudoparticles expressing the spike protein of several SARS-CoV-2 variants, we evaluated the magnitude and breadth of the neutralizing antibody response over time in individuals after infection and in mRNA-vaccinated individuals. We observed that boosting increases the magnitude of the antibody response to wild-type (D614), Beta, Delta, and Omicron variants; however, the Omicron variant was the most resistant to neutralization. We further observed that vaccinated healthy adults had robust and broad antibody responses, whereas responses may have been reduced in vaccinated pregnant women, underscoring the importance of learning how to maximize mRNA vaccine responses in pregnant populations. Findings from this study show substantial heterogeneity in the magnitude and breadth of responses after infection and mRNA vaccination and may support the addition of more conserved viral antigens to existing SARS-CoV-2 vaccines.

A Review of Single-Source Deep Unsupervised Visual Domain Adaptation.

Large-scale labeled training datasets have enabled deep neural networks to excel across a wide range of benchmark vision tasks. However, in many applications, it is prohibitively expensive and time-consuming to obtain large quantities of labeled data. To cope with limited labeled training data, many have attempted to directly apply models trained on a large-scale labeled source domain to another sparsely labeled or unlabeled target domain. Unfortunately, direct transfer across domains often performs poorly due to the presence of domain shift or dataset bias. Domain adaptation (DA) is a machine learning paradigm that aims to learn a model from a source domain that can perform well on a different (but related) target domain. In this article, we review the latest single-source deep unsupervised DA methods focused on visual tasks and discuss new perspectives for future research. We begin with the definitions of different DA strategies and the descriptions of existing benchmark datasets. We then summarize and compare different categories of single-source unsupervised DA methods, including discrepancy-based methods, adversarial discriminative methods, adversarial generative methods, and self-supervision-based methods. Finally, we discuss future research directions with challenges and possible solutions.

Practice of Telehealth in Otolaryngology: A Scoping Review in the Era of COVID-19.

OBJECTIVE: The COVID-19 pandemic has spurred widespread adoption and advancement in telehealth activities, representing a marked change in otolaryngology practice patterns. The present study undertakes a scoping review of research focused on telehealth in otolaryngology (teleotolaryngology) to identify key themes and commonly utilized outcome measures that will assist future development in this growing field. DATA SOURCES: PubMed, Embase, and Cochrane databases and reference review. REVIEW METHODS: Per guidelines of the PRISMA Extension for Scoping Reviews, we performed database queries using a comprehensive search strategy developed in collaboration with research librarians at the Columbia University Irving Medical Center. We identified 596 unique references to undergo title and abstract review by 2 independent reviewers, leaving 439 studies for full-text review. RESULTS: We included 285 studies for extraction of notable findings, leaving 262 unique studies after accounting for content overlap. We identified core outcome measures, including patient and provider satisfaction, costs and benefits, quality of care, feasibility, and access to care. Publication volume increased markedly over time, though only 4% of studies incorporated randomized study group assignment. Using an iterative approach to thematic development, we organized article content across 5 main themes: (1) exploration of teleotolaryngology evolution, (2) role in virtual clinical encounters, (3) applications in interdisciplinary care and educational initiatives, (4) emerging and innovative technologies, and (5) barriers to implementation. CONCLUSION: This scoping review of teleotolaryngology documents its evolution and identifies current use cases, limitations, and emerging applications, providing a foundation from which to build future studies, inform policy decision making, and facilitate implementation where appropriate.

Crystal structure of silver pentazolates AgN5 and AgN6.

Silver pentazolate, a high energy density compound containing the cyclo-N5- anion, has recently been synthesized under ambient conditions. However, due to high sensitivity to irradiation, its crystal structure has not been determined. In this work, silver-nitrogen crystalline compounds under ambient conditions and at high pressures, up to 100 GPa, are predicted and characterized by performing first-principles evolutionary crystal structure searching with variable stoichiometry. It is found that newly discovered AgN5 and AgN6 are the only thermodynamically stable silver-nitrogen compounds at pressures between 42 and 80 GPa. In contrast to AgN5, the pentazolate AgN6 compound contains N2 diatomic molecules in addition to cyclo-N5-. These AgN5 and AgN6 crystals are metastable under ambient conditions with positive formation enthalpies of 54.95 kJ mol-1 and 46.24 kJ mol-1, respectively. The underlying cause of the stability of cyclo-N5- silver pentazolates is the enhanced aromaticity enabled by the charge transfer from silver atoms to nitrogen rings. To aid in the experimental identification of these materials, calculated Raman spectra are reported at ambient pressure: the frequencies of N5- vibrational modes of AgN5 are in good agreement with those measured in the experiment.

Assessment of Allergic and Anaphylactic Reactions to mRNA COVID-19 Vaccines With Confirmatory Testing in a US Regional Health System.

Importance: As of May 2021, more than 32 million cases of COVID-19 have been confirmed in the United States, resulting in more than 615 000 deaths. Anaphylactic reactions associated with the Food and Drug Administration (FDA)-authorized mRNA COVID-19 vaccines have been reported. Objective: To characterize the immunologic mechanisms underlying allergic reactions to these vaccines. Design, Setting, and Participants: This case series included 22 patients with suspected allergic reactions to mRNA COVID-19 vaccines between December 18, 2020, and January 27, 2021, at a large regional health care network. Participants were individuals who received at least 1 of the following International Statistical Classification of Diseases and Related Health Problems, Tenth Revision anaphylaxis codes: T78.2XXA, T80.52XA, T78.2XXD, or E949.9, with documentation of COVID-19 vaccination. Suspected allergy cases were identified and invited for follow-up allergy testing. Exposures: FDA-authorized mRNA COVID-19 vaccines. Main Outcomes and Measures: Allergic reactions were graded using standard definitions, including Brighton criteria. Skin prick testing was conducted to polyethylene glycol (PEG) and polysorbate 80 (P80). Histamine (1 mg/mL) and filtered saline (negative control) were used for internal validation. Basophil activation testing after stimulation for 30 minutes at 37 °C was also conducted. Concentrations of immunoglobulin (Ig) G and IgE antibodies to PEG were obtained to determine possible mechanisms. Results: Of 22 patients (20 [91%] women; mean [SD] age, 40.9 [10.3] years; 15 [68%] with clinical allergy history), 17 (77%) met Brighton anaphylaxis criteria. All reactions fully resolved. Of patients who underwent skin prick tests, 0 of 11 tested positive to PEG, 0 of 11 tested positive to P80, and 1 of 10 (10%) tested positive to the same brand of mRNA vaccine used to vaccinate that individual. Among these same participants, 10 of 11 (91%) had positive basophil activation test results to PEG and 11 of 11 (100%) had positive basophil activation test results to their administered mRNA vaccine. No PEG IgE was detected; instead, PEG IgG was found in tested individuals who had an allergy to the vaccine. Conclusions and Relevance: Based on this case series, women and those with a history of allergic reactions appear at have an elevated risk of mRNA vaccine allergy. Immunological testing suggests non-IgE-mediated immune responses to PEG may be responsible in most individuals.

Performance of a Convolutional Neural Network and Explainability Technique for 12-Lead Electrocardiogram Interpretation.

Importance: Millions of clinicians rely daily on automated preliminary electrocardiogram (ECG) interpretation. Critical comparisons of machine learning-based automated analysis against clinically accepted standards of care are lacking. Objective: To use readily available 12-lead ECG data to train and apply an explainability technique to a convolutional neural network (CNN) that achieves high performance against clinical standards of care. Design, Setting, and Participants: This cross-sectional study was conducted using data from January 1, 2003, to December 31, 2018. Data were obtained in a commonly available 12-lead ECG format from a single-center tertiary care institution. All patients aged 18 years or older who received ECGs at the University of California, San Francisco, were included, yielding a total of 365 009 patients. Data were analyzed from January 1, 2019, to March 2, 2021. Exposures: A CNN was trained to predict the presence of 38 diagnostic classes in 5 categories from 12-lead ECG data. A CNN explainability technique called LIME (Linear Interpretable Model-Agnostic Explanations) was used to visualize ECG segments contributing to CNN diagnoses. Main Outcomes and Measures: Area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were calculated for the CNN in the holdout test data set against cardiologist clinical diagnoses. For a second validation, 3 electrophysiologists provided consensus committee diagnoses against which the CNN, cardiologist clinical diagnosis, and MUSE (GE Healthcare) automated analysis performance was compared using the F1 score; AUC, sensitivity, and specificity were also calculated for the CNN against the consensus committee. Results: A total of 992 748 ECGs from 365 009 adult patients (mean [SD] age, 56.2 [17.6] years; 183 600 women [50.3%]; and 175 277 White patients [48.0%]) were included in the analysis. In 91 440 test data set ECGs, the CNN demonstrated an AUC of at least 0.960 for 32 of 38 classes (84.2%). Against the consensus committee diagnoses, the CNN had higher frequency-weighted mean F1 scores than both cardiologists and MUSE in all 5 categories (CNN frequency-weighted F1 score for rhythm, 0.812; conduction, 0.729; chamber diagnosis, 0.598; infarct, 0.674; and other diagnosis, 0.875). For 32 of 38 classes (84.2%), the CNN had AUCs of at least 0.910 and demonstrated comparable F1 scores and higher sensitivity than cardiologists, except for atrial fibrillation (CNN F1 score, 0.847 vs cardiologist F1 score, 0.881), junctional rhythm (0.526 vs 0.727), premature ventricular complex (0.786 vs 0.800), and Wolff-Parkinson-White (0.800 vs 0.842). Compared with MUSE, the CNN had higher F1 scores for all classes except supraventricular tachycardia (CNN F1 score, 0.696 vs MUSE F1 score, 0.714). The LIME technique highlighted physiologically relevant ECG segments. Conclusions and Relevance: The results of this cross-sectional study suggest that readily available ECG data can be used to train a CNN algorithm to achieve comparable performance to clinical cardiologists and exceed the performance of MUSE automated analysis for most diagnoses, with some exceptions. The LIME explainability technique applied to CNNs highlights physiologically relevant ECG segments that contribute to the CNN's diagnoses.

Illuminating the Fc dependence of SARS-CoV-2 neutralization.

Monoclonal antibodies show efficacy in treating COVID-19, but the functional requirements for protection are unclear. In this issue of Immunity, Ullah et al. (2021) develop a stable SARS-CoV-2 reporter virus and use bioluminescence imaging to longitudinally monitor infection and assess neutralizing monoclonal antibody interventions in mice. They find that antibody-mediated protection depends on the Fc domain and Fc-gamma receptor-expressing immune cells.