Potential of conserved antigenic sites in development of universal SARS-like coronavirus vaccines

Given pandemic risks of zoonotic SARS-CoV-2 variants and other SARS-like coronaviruses in the future, it is valuable to perform studies on conserved antigenic sites to design universal SARS-like coronavirus vaccines. By using antibodies obtained from convalescent COVID-19 patients, we succeeded in functional comparison of conserved antigenic sites at multiple aspects with each other, and even with SARS-CoV-2 unique antigenic sites, which promotes the cognition of process of humoral immune response to the conserved antigenic sites. The conserved antigenic sites between SARS-CoV-2 and SARS-CoV can effectively induce affinity maturation of cross-binding antibodies, finally resulting in broadly neutralizing antibodies against multiple variants of concern, which provides an important basis for universal vaccine design, however they are subdominant, putatively due to their lower accessibility relative to SARS-CoV-2 unique antigenic sites. Furthermore, we preliminarily design RBDs to improve the immunogenicity of these conserved antigenic sites. Our study focusing on conserved antigenic sites provides insights for promoting the development of universal SARS-like coronavirus vaccines, thereby enhancing our pandemic preparedness.

Machine Learning for Encrypted Malicious Traffic Detection: Approaches, Datasets and Comparative Study (preprint)

As people's demand for personal privacy and data security becomes a priority, encrypted traffic has become mainstream in the cyber world. However, traffic encryption is also shielding malicious and illegal traffic introduced by adversaries, from being detected. This is especially so in the post-COVID-19 environment where malicious traffic encryption is growing rapidly. Common security solutions that rely on plain payload content analysis such as deep packet inspection are rendered useless. Thus, machine learning based approaches have become an important direction for encrypted malicious traffic detection. In this paper, we formulate a universal framework of machine learning based encrypted malicious traffic detection techniques and provided a systematic review. Furthermore, current research adopts different datasets to train their models due to the lack of well-recognized datasets and feature sets. As a result, their model performance cannot be compared and analyzed reliably. Therefore, in this paper, we analyse, process and combine datasets from 5 different sources to generate a comprehensive and fair dataset to aid future research in this field. On this basis, we also implement and compare 10 encrypted malicious traffic detection algorithms. We then discuss challenges and propose future directions of research.

Economic Recession Prediction Using Deep Neural Network (preprint)

We investigate the effectiveness of different machine learning methodologies in predicting economic cycles. We identify the deep learning methodology of Bi-LSTM with Autoencoder as the most accurate model to forecast the beginning and end of economic recessions in the U.S. We adopt commonly-available macro and market-condition features to compare the ability of different machine learning models to generate good predictions both in-sample and out-of-sample. The proposed model is flexible and dynamic when both predictive variables and model coefficients vary over time. It provided good out-of-sample predictions for the past two recessions and early warning about the COVID-19 recession.

Distinguish Coronavirus Disease 2019 Patients in General Surgery Emergency by CIAAD Scale: Development and Validation of a Prediction Model Based on 822 Cases in China (preprint)

IMPORTANCE In the epidemic, surgeons cannot distinguish infectious acute abdomen patients suspected COVID-19 quickly and effectively. OBJECTIVE To develop and validate a predication model, presented as nomogram and scale, to distinguish infectious acute abdomen patients suspected coronavirus disease 2019 (COVID-19). DESIGN Diagnostic model based on retrospective case series. SETTING Two hospitals in Wuhan and Beijing, China. PTRTICIPANTS 584 patients admitted to hospital with laboratory confirmed SARS-CoV-2 from 2 Jan 2020 to15 Feb 2020 and 238 infectious acute abdomen patients receiving emergency operation from 28 Feb 2019 to 3 Apr 2020. METHODS LASSO regression and multivariable logistic regression analysis were conducted to develop the prediction model in training cohort. The performance of the nomogram was evaluated by calibration curves, receiver operating characteristic (ROC) curves, decision curve analysis (DCA) and clinical impact curves in training and validation cohort. A simplified screening scale and managing algorithm was generated according to the nomogram. RESULTS Six potential COVID-19 prediction variables were selected and the variable abdominal pain was excluded for overmuch weight. The five potential predictors, including fever, chest computed tomography (CT), leukocytes (white blood cells, WBC), C-reactive protein (CRP) and procalcitonin (PCT), were all independent predictors in multivariable logistic regression analysis (p[≤]0.001) and the nomogram, named COVID-19 Infectious Acute Abdomen Distinguishment (CIAAD) nomogram, was generated. The CIAAD nomogram showed good discrimination and calibration (C-index of 0.981 (95% CI, 0.963 to 0.999) and AUC of 0.970 (95% CI, 0.961 to 0.982)), which was validated in the validation cohort (C-index of 0.966 (95% CI, 0.960 to 0.972) and AUC of 0.966 (95% CI, 0.957 to 0.975)). Decision curve analysis revealed that the CIAAD nomogram was clinically useful. The nomogram was further simplified into the CIAAD scale. CONCLUSIONS We established an easy and effective screening model and scale for surgeons in emergency department to distinguish COVID-19 patients from infectious acute abdomen patients. The algorithm based on CIAAD scale will help surgeons manage infectious acute abdomen patients suspected COVID-19 more efficiently.