COVIDx CXR-4: An Expanded Multi-Institutional Open-Source Benchmark Dataset for Chest X-ray Image-Based Computer-Aided COVID-19 Diagnostics (preprint)

The global ramifications of the COVID-19 pandemic remain significant, exerting persistent pressure on nations even three years after its initial outbreak. Deep learning models have shown promise in improving COVID-19 diagnostics but require diverse and larger-scale datasets to improve performance. In this paper, we introduce COVIDx CXR-4, an expanded multi-institutional open-source benchmark dataset for chest X-ray image-based computer-aided COVID-19 diagnostics. COVIDx CXR-4 expands significantly on the previous COVIDx CXR-3 dataset by increasing the total patient cohort size by greater than 2.66 times, resulting in 84,818 images from 45,342 patients across multiple institutions. We provide extensive analysis on the diversity of the patient demographic, imaging metadata, and disease distributions to highlight potential dataset biases. To the best of the authors' knowledge, COVIDx CXR-4 is the largest and most diverse open-source COVID-19 CXR dataset and is made publicly available as part of an open initiative to advance research to aid clinicians against the COVID-19 disease.

The effect of COVID-19 restrictions on atmospheric new particle formation in Beijing

During the COVID-19 lockdown, the dramatic reduction of anthropogenic emissions provided a unique opportunity to investigate the effects of reduced anthropogenic activity and primary emissions on atmospheric chemical processes and the consequent formation of secondary pollutants. Here, we utilize comprehensive observations to examine the response of atmospheric new particle formation (NPF) to the changes in the atmospheric chemical cocktail. We find that the main clustering process was unaffected by the drastically reduced traffic emissions, and the formation rate of 1.5 nm particles remained unaltered. However, particle survival probability was enhanced due to an increased particle growth rate (GR) during the lockdown period, explaining the enhanced NPF activity in earlier studies. For GR at 1.5–3 nm, sulfuric acid (SA) was the main contributor at high temperatures, whilst there were unaccounted contributing vapors at low temperatures. For GR at 3–7 and 7–15 nm, oxygenated organic molecules (OOMs) played a major role. Surprisingly, OOM composition and volatility were insensitive to the large change of atmospheric NOx concentration;instead the associated high particle growth rates and high OOM concentration during the lockdown period were mostly caused by the enhanced atmospheric oxidative capacity. Overall, our findings suggest a limited role of traffic emissions in NPF.

A global systematic analysis of the occurrence, severity, and recovery pattern of long COVID in 2020 and 2021 (preprint)

ImportanceWhile much of the attention on the COVID-19 pandemic was directed at the daily counts of cases and those with serious disease overwhelming health services, increasingly, reports have appeared of people who experience debilitating symptoms after the initial infection. This is popularly known as long COVID. ObjectiveTo estimate by country and territory of the number of patients affected by long COVID in 2020 and 2021, the severity of their symptoms and expected pattern of recovery DesignWe jointly analyzed ten ongoing cohort studies in ten countries for the occurrence of three major symptom clusters of long COVID among representative COVID cases. The defining symptoms of the three clusters (fatigue, cognitive problems, and shortness of breath) are explicitly mentioned in the WHO clinical case definition. For incidence of long COVID, we adopted the minimum duration after infection of three months from the WHO case definition. We pooled data from the contributing studies, two large medical record databases in the United States, and findings from 44 published studies using a Bayesian meta-regression tool. We separately estimated occurrence and pattern of recovery in patients with milder acute infections and those hospitalized. We estimated the incidence and prevalence of long COVID globally and by country in 2020 and 2021 as well as the severity-weighted prevalence using disability weights from the Global Burden of Disease study. ResultsAnalyses are based on detailed information for 1906 community infections and 10526 hospitalized patients from the ten collaborating cohorts, three of which included children. We added published data on 37262 community infections and 9540 hospitalized patients as well as ICD-coded medical record data concerning 1.3 million infections. Globally, in 2020 and 2021, 144.7 million (95% uncertainty interval [UI] 54.8-312.9) people suffered from any of the three symptom clusters of long COVID. This corresponds to 3.69% (1.38-7.96) of all infections. The fatigue, respiratory, and cognitive clusters occurred in 51.0% (16.9-92.4), 60.4% (18.9-89.1), and 35.4% (9.4-75.1) of long COVID cases, respectively. Those with milder acute COVID-19 cases had a quicker estimated recovery (median duration 3.99 months [IQR 3.84-4.20]) than those admitted for the acute infection (median duration 8.84 months [IQR 8.10-9.78]). At twelve months, 15.1% (10.3-21.1) continued to experience long COVID symptoms. Conclusions and relevanceThe occurrence of debilitating ongoing symptoms of COVID-19 is common. Knowing how many people are affected, and for how long, is important to plan for rehabilitative services and support to return to social activities, places of learning, and the workplace when symptoms start to wane. Key PointsO_ST_ABSQuestionC_ST_ABSWhat are the extent and nature of the most common long COVID symptoms by country in 2020 and 2021? FindingsGlobally, 144.7 million people experienced one or more of three symptom clusters (fatigue; cognitive problems; and ongoing respiratory problems) of long COVID three months after infection, in 2020 and 2021. Most cases arose from milder infections. At 12 months after infection, 15.1% of these cases had not yet recovered. MeaningThe substantial number of people with long COVID are in need of rehabilitative care and support to transition back into the workplace or education when symptoms start to wane.

BridgeDPI: A Novel Graph Neural Network for Predicting Drug-Protein Interactions (preprint)

Motivation: Exploring drug-protein interactions (DPIs) work as a pivotal step in drug discovery. The fast expansion of available biological data enables computational methods effectively assist in experimental methods. Among them, deep learning methods extract features only from basic characteristics, such as protein sequences, molecule structures. Others achieve significant improvement by learning from not only sequences/molecules but the protein-protein and drug-drug associations (PPAs and DDAs). The PPAs and DDAs are generally obtained by using computational methods. However, existing computational methods have some limitations, resulting in low-quality PPAs and DDAs that hamper the prediction performance. Therefore, we hope to develop a novel supervised learning method to learn the PPAs and DDAs effectively and thereby improve the prediction performance of the specific task of DPI. Results: In this research, we propose a novel deep learning framework, namely BridgeDPI. BridgeDPI introduces a class of nodes named hyper-nodes, which bridge different proteins/drugs to work as PPAs and DDAs. The hyper-nodes can be supervised learned for the specific task of DPI since the whole process is an end-to-end learning. Consequently, such a model would improve prediction performance of DPI. In three real-world datasets, we further demonstrate that BridgeDPI outperforms state-of-the-art methods. Moreover, ablation studies verify the effectiveness of the hyper-nodes. Last, in an independent verification, BridgeDPI explores the candidate bindings among COVID-19's proteins and various antiviral drugs. And the predictive results accord with the statement of the World Health Organization and Food and Drug Administration, showing the validity and reliability of BridgeDPI.

A novel viral protein translation mechanism reveals mitochondria as a target for antiviral drug development (preprint)

The ongoing Severe Acute Respiratory Syndrome coronavirus 2 (SARS-CoV-2) pandemic has acutely highlighted the need to identify new treatment strategies for viral infections. Here we present a pivotal molecular mechanism of viral protein translation that relies on the mitochondrial translation machinery. We found that rare codons such as Leu-TTA are highly enriched in many viruses, including SARS-CoV-2, and these codons are essential for the regulation of viral protein expression. SARS-CoV-2 controls the translation of its spike gene by hijacking host mitochondria through 5' leader and 3'UTR sequences that contain mitochondrial localization signals and activate the EGR1 pathway. Mitochondrial-targeted drugs such as lonidamine and polydatin significantly repress rare codon-driven gene expression and viral replication. This study identifies an unreported viral protein translation mechanism and opens up a novel avenue for developing antiviral drugs.