Virus surveillance and genomic analyses to identify emerging new virus variants. recommended for the 2025–2026 vaccine formula has been replaced. Real-time ...

Through the study of viral genomes, scientists can identify mutations that might influence transmissibility, immune escape, and the effectiveness of vaccines.

This research examines how aging and inflammatory processes impair the development of SARS-CoV-2-specific immune responses. The study investigates the mechanisms by which advancing age and chronic inflammation compromise the body's ability to generate effective adaptive immunity against the coronavirus. Understanding these immunological limitations in older populations is critical for improving vaccine efficacy and therapeutic strategies in elderly individuals who face elevated disease severity risks. The findings have significant implications for public health interventions targeting vulnerable demographics and inform vaccine design considerations for aging populations.

ViralMultiNet represents a significant advancement in viral protein functional annotation, addressing critical limitations of sequence-only models through a structure-aware multimodal framework. The system integrates multi-scale k-mer encodings with semantic embeddings from UniProt annotations, leveraging gated fusion and knowledge distillation techniques. Validated on 66,011 SARS-CoV-2 wastewater metagenomics samples, the framework achieved a macro F1 score of 0.921 and AUC of 0.983 while reducing computational time by 40.4%. The distilled student model maintained comparable performance with negligible degradation. Notably, attention analysis demonstrated precise alignment with experimentally validated functional domains, including the receptor-binding domain and S1/S2 cleavage site. This interpretable, scalable solution effectively generalizes across emerging variants, positioning it as a valuable tool for wastewater-based surveillance systems and pandemic preparedness initiatives.

# Refined Phenotyping of Vaccine Responses Reveals Transcriptomic Determinants of Neutralizing Antibody Heterogeneity This research employs advanced phenotyping techniques to investigate the molecular mechanisms underlying variable immune responses to vaccines. By analyzing transcriptomic profiles, the study identifies genetic and molecular factors that explain why individuals develop different levels of neutralizing antibodies following vaccination. These findings provide critical insights into immune heterogeneity, potentially enabling personalized vaccination strategies and improved vaccine design. Understanding these transcriptomic determinants could facilitate identification of individuals with suboptimal immune responses, allowing for targeted interventions to enhance protective immunity and public health outcomes.

This systematic review examines neurological manifestations of post-viral syndromes, particularly Long COVID, across adult and pediatric populations. Analyzing 23 studies published between January 2020 and May 2025, the research identifies persistent fatigue, cognitive dysfunction, headaches, sleep disturbances, dysautonomia, and neuropathic symptoms as predominant neurological features. The study reveals a multifactorial pathophysiology involving immune dysregulation, neuroinflammation, endothelial dysfunction, mitochondrial impairment, and autonomic nervous system disruption. Pediatric cases demonstrate similar symptom profiles with notable cognitive and behavioral impacts. These findings underscore neurological dysfunction as a fundamental component of Long COVID, establishing the complexity of underlying mechanisms and informing future therapeutic strategies for this emerging clinical condition.

Machine learning is revolutionizing protein dynamics characterization by overcoming limitations of traditional structural biology techniques. While X-ray crystallography and cryo-electron microscopy provide valuable static snapshots, they fail to capture the conformational fluctuations essential to biological function. Computational methods like molecular dynamics offer dynamic insights but are computationally expensive and face sampling limitations. Machine learning addresses these challenges by refining force fields, optimizing adaptive sampling strategies, and automating cryo-EM and NMR workflows. Additionally, deep learning and protein language models enhance structure prediction capabilities. By integrating physical principles with data-driven approaches, machine learning bridges the gap between static structural representations and dynamic functional ensembles, significantly advancing structural biology and drug discovery applications.

This research article examines Long COVID through a network disorder framework, proposing mechanism-anchored approaches to understanding the condition's complex pathophysiology. Rather than treating Long COVID as a single disease entity, the authors analyze it as a multisystem disorder involving interconnected biological networks. The study integrates evidence from immunological, neurological, and vascular mechanisms to explain persistent symptoms following acute COVID-19 infection. By anchoring explanations in specific biological mechanisms, the research provides a comprehensive model for understanding disease heterogeneity among Long COVID patients. This network-based perspective offers important insights for developing targeted therapeutic interventions and improving clinical management strategies for this emerging chronic condition.

# Summary Recent research has identified a potential autoimmune mechanism underlying Long COVID symptoms, a condition affecting numerous SARS-CoV-2 survivors. Patients experience persistent manifestations including debilitating fatigue, cognitive dysfunction, cardiac palpitations, and musculoskeletal pain that can persist for extended periods following acute infection. This study advances understanding of Long COVID's pathophysiology by pointing toward autoimmune dysfunction as a key driver of these prolonged symptoms. Such findings have significant implications for developing targeted therapeutic interventions and improving patient outcomes for the millions experiencing Long COVID complications.

This study examines clinical characteristics and antibody responses among pregnant women infected with Omicron variants in China from December 2022 to April 2023. Researchers investigated 223 pregnant women with at least six weeks gestation and 31 healthy controls to evaluate humoral immune responses following natural SARS-CoV-2 infection, particularly in those previously vaccinated with inactivated vaccines. The research addresses critical knowledge gaps regarding how immune imprinting from inactivated vaccines may influence cross-neutralization against emerging Omicron subvariants. This cross-sectional investigation provides important insights into pregnancy-specific immune dynamics during COVID-19 infection and vaccine responses, contributing valuable data for understanding maternal and fetal health implications during the pandemic.

HLA-C is a major histocompatibility complex class I gene in humans, identified as Gene ID 3107 in the National Center for Biotechnology Information database. This gene encodes a protein involved in the immune system's ability to present antigens to T cells, playing a crucial role in immune recognition and response. Located within the human genome, HLA-C is essential for transplant compatibility assessment, disease susceptibility studies, and understanding immune-mediated conditions. The NCBI database provides comprehensive genomic data, including sequencing information, expression profiles, and mapped phenotypes associated with HLA-C variants, supporting clinical and research applications in immunology and precision medicine.

This study presents an innovative computational framework for optimizing therapeutic mRNA sequences by combining deep learning with evolutionary algorithms. Researchers employ a pretrained CodonTransformer to generate initial mRNA candidates, followed by a genetic algorithm that iteratively refines sequences through codon-aware modifications. The optimization targets multiple critical parameters: translation efficiency metrics including codon adaptation index and tRNA adaptation index, structural stability through mRNA secondary structure analysis, and reduced immunogenic potential by minimizing problematic motif frequencies. Results demonstrate substantial improvements across successive generations, achieving enhanced codon adaptation and tRNA matching while maintaining stable secondary structures and low immunogenicity. This integrated approach offers a promising alternative to existing mRNA design methodologies for developing safer and more effective therapeutic vaccines and treatments.

Artificial intelligence-directed computational protein design has revolutionized therapeutic discovery, particularly for vaccines and antibody-based treatments against infectious diseases. The COVID-19 pandemic demonstrated both the transformative potential and significant limitations of these technologies. AI-driven approaches successfully accelerated viral characterization, vaccine development, and broadly neutralizing antibody design. However, critical challenges emerged, including data bias, model interpretability issues, experimental validation bottlenecks, and regulatory framework integration gaps. The pandemic revealed a crucial disconnect between computational promise and translational readiness, necessitating closer collaboration between computational design, laboratory experimentation, and clinical evaluation. Additionally, rapid AI innovation has outpaced established regulatory pathways, highlighting the need for enhanced oversight mechanisms before these technologies can serve as reliable pandemic preparedness cornerstones.

This review examines the application of machine learning and artificial intelligence in predicting immunogenicity risks associated with biologic therapies during preclinical development. Anti-drug antibodies resulting from unwanted immune responses pose significant challenges to clinical development and patient outcomes. Currently, predicting immunogenicity before clinical trials remains impossible due to immune system complexity and multiple contributing factors. However, advances in computational methods offer promising opportunities for improved prediction accuracy. Industry experts and academics convened at an EMBL-EBI workshop to assess AI and ML contributions to immunogenicity prediction, review current industry practices, and identify future opportunities for deploying these transformative technologies in therapeutic development.

People vaccinated against COVID in Germany in 2023 experienced reduced hospitalization rates, fewer long-COVID diagnoses, and lower all-cause mortality than those who weren't vaccinated, according to a new study published in Eurosurveillance. These clinical benefits translated to significant reductions in health care costs and lost productivity due to sick leave.

The World Health Organization is monitoring BA.3.2, a new COVID-19 variant circulating across the United States as warmer temperatures approach. This variant, a descendant of Omicron, has been detected in 31 states and raises concerns about its potential to evade immunity from prior vaccination or infection due to modifications in its spike protein. Dr. Sonia Navas-Martin, a prominent microbiology and immunology professor at Drexel University College of Medicine, provides expert analysis on the variant's characteristics and public health implications. Her research focuses on immune responses to viral infections and the roles of spike proteins in coronavirus transmission and disease development, offering valuable insights into whether this emerging variant warrants heightened concern.

A heavily mutated COVID-19 variant designated BA.3.2, colloquially known as "Cicada," is spreading across over thirty U.S. states and gaining global traction. This variant emerged over a year ago but accelerated significantly in late 2024. BA.3.2 is distinguished by substantial genetic mutations in its spike protein, potentially enabling it to evade immunity acquired through vaccination or prior infection. Research published in the CDC's Morbidity and Mortality Weekly Report suggests these genetic modifications may reduce protective immunity. Consequently, public health officials are closely monitoring this "hyper-mutated" strain, which the World Health Organization designated as a variant under monitoring in December 2025. While overall COVID infections are declining nationally, BA.3.2's unique characteristics warrant continued epidemiological surveillance.

The Vaccine Adverse Event Reporting System (VAERS) serves as the nation's primary early warning system for monitoring the safety of FDA-approved vaccines and those authorized for public health emergencies. Co-managed by the CDC and FDA, VAERS accepts and analyzes reports of potential adverse events following vaccination from patients, healthcare providers, family members, and manufacturers. While healthcare providers and manufacturers are legally obligated to report certain events, submission of a report does not establish causation between a vaccine and an adverse event. When patterns suggest possible vaccine-associated health problems, the CDC and FDA conduct further investigation and implement necessary interventions. This comprehensive surveillance system enables continuous safety monitoring across the vaccinated population.

The European Centre for Disease Prevention and Control (ECDC) maintains a systematic approach to monitoring and classifying SARS-CoV-2 variants of concern as of April 24, 2026. The SAVE Working Group, comprising multidisciplinary experts in virology, bioinformatics, and epidemiology, convenes monthly to assess emerging variants' epidemiological impact across the EU/EEA and globally. ECDC employs three classification categories—variant under monitoring, variant of interest, and variant of concern—to communicate escalating risk levels. Evidence is continuously evaluated through epidemic intelligence and genomic surveillance, with classifications updated accordingly. Detailed surveillance data, including variant distribution and country-specific epidemiological updates, are published through the European Respiratory Virus Surveillance Summary, supporting coordinated public health response efforts.

A comprehensive safety evaluation of the mRNA-1273 COVID-19 vaccine, published in Human Vaccines & Immunotherapeutics, confirms a strong and well-characterized safety profile from clinical development through widespread global deployment. The review analyzed clinical trial data, pharmacovigilance findings, and real-world observational studies conducted across tens of thousands of participants. Preclinical studies demonstrated strong immunogenicity with acceptable tolerability, while clinical trials revealed a predictable, manageable reactogenicity profile featuring common mild-to-moderate transient adverse events such as injection-site pain, fatigue, headache, and myalgia. Serious adverse events occurred at comparable rates between vaccine and placebo groups. Post-authorization surveillance through global pharmacovigilance systems, including the Vaccine Adverse Event Reporting System and international regulatory databases, continued to validate the vaccine's safety profile during unprecedented global use.