The global spread of SARS-CoV-2 is posing major public health challenges. One feature of SARS-CoV-2 spike protein is the insertion of multi-basic residues at the S1/S2 subunit cleavage site. Here, we find that the virus with intact spike (Sfull) preferentially enters cells via fusion at the plasma membrane, whereas a clone (Sdel) with deletion disrupting the multi-basic S1/S2 site utilizes an endosomal entry pathway. Using Sdel as model, we perform a genome-wide CRISPR screen and identify several endosomal entry-specific regulators. Experimental validation of hits from the CRISPR screen shows that host factors regulating the surface expression of angiotensin-converting enzyme 2 (ACE2) affect entry of Sfull virus. Animal-to-animal transmission with the Sdel virus is reduced compared to Sfull in the hamster model. These findings highlight the critical role of the S1/S2 boundary of SARS-CoV-2 spike protein in modulating virus entry and transmission and provide insights into entry of coronaviruses.

For offering proactive services (e.g., personalized exercise recommendation) to the students in computer supported intelligent education, one of the fundamental tasks is predicting student performance (e.g., scores) on future exercises, where it is necessary to track the change of each student's knowledge acquisition during her exercising activities. Unfortunately, to the best of our knowledge, existing approaches can only exploit the exercising records of students, and the problem of extracting rich information existed in the materials (e.g., knowledge concepts, exercise content) of exercises to achieve both more precise prediction of student performance and more interpretable analysis of knowledge acquisition remains underexplored. To this end, in this paper, we present a holistic study of student performance prediction. To directly achieve the primary goal of performance prediction, we first propose a general Exercise-Enhanced Recurrent Neural Network (EERNN) framework by exploring both student's exercising records and the text content of corresponding exercises. In EERNN, we simply summarize each student's state into an integrated vector and trace it with a recurrent neural network, where we design a bidirectional LSTM to learn the encoding of each exercise from its content. For making final predictions, we design two implementations on the basis of EERNN with different prediction strategies, i.e., EERNNM with Markov property and EERNNA with Attention mechanism. Then, to explicitly track student's knowledge acquisition on multiple knowledge concepts, we extend EERNN to an explainable Exercise-aware K nowledge T racing (EKT) framework by incorporating the knowledge concept information, where the student's integrated state vector is now extended to a knowledge state matrix. In EKT, we further develop a memory network for quantifying how much each exercise can affect the mastery of students on multiple knowledge concepts during the exercising process. Finally, we conduct extensive experiments and evaluate both EERNN and EKT frameworks on a large-scale real-world data. The results in both general and cold-start scenarios clearly demonstrate the effectiveness of two frameworks in student performance prediction as well as the superior interpretability of EKT.

Lack of detailed knowledge of SARS-CoV-2 infection has been hampering the development of treatments for coronavirus disease 2019 (COVID-19). Here, we report that RNA triggers the liquid–liquid phase separation (LLPS) of the SARS-CoV-2 nucleocapsid protein, N. By analyzing all 29 proteins of SARS-CoV-2, we find that only N is predicted as an LLPS protein. We further confirm the LLPS of N during SARS-CoV-2 infection. Among the 100,849 genome variants of SARS-CoV-2 in the GISAID database, we identify that ~37% (36,941) of the genomes contain a specific trio-nucleotide polymorphism (GGG-to-AAC) in the coding sequence of N, which leads to the amino acid substitutions, R203K/G204R. Interestingly, NR203K/G204R exhibits a higher propensity to undergo LLPS and a greater effect on IFN inhibition. By screening the chemicals known to interfere with N-RNA binding in other viruses, we find that (-)-gallocatechin gallate (GCG), a polyphenol from green tea, disrupts the LLPS of N and inhibits SARS-CoV-2 replication. Thus, our study reveals that targeting N-RNA condensation with GCG could be a potential treatment for COVID-19.

Cognitive diagnosis is a fundamental issue in intelligent education, which aims to discover the proficiency level of students on specific knowledge concepts. Existing approaches usually mine linear interactions of student exercising process by manual-designed function (e.g., logistic function), which is not sufficient for capturing complex relations between students and exercises. In this paper, we propose a general Neural Cognitive Diagnosis (NeuralCD) framework, which incorporates neural networks to learn the complex exercising interactions, for getting both accurate and interpretable diagnosis results. Specifically, we project students and exercises to factor vectors and leverage multi neural layers for modeling their interactions, where the monotonicity assumption is applied to ensure the interpretability of both factors. Furthermore, we propose two implementations of NeuralCD by specializing the required concepts of each exercise, i.e., the NeuralCDM with traditional Q-matrix and the improved NeuralCDM+ exploring the rich text content. Extensive experimental results on real-world datasets show the effectiveness of NeuralCD framework with both accuracy and interpretability.

The global spread of SARS-CoV-2 is posing major public health challenges. One unique feature of SARS-CoV-2 spike protein is the insertion of multi-basic residues at the S1/S2 subunit cleavage site, the function of which remains uncertain. We found that the virus with intact spike (Sfull) preferentially enters cells via fusion at the plasma membrane, whereas a clone (Sdel) with deletion disrupting the multi-basic S1/S2 site instead utilizes a less efficient endosomal entry pathway. This idea was supported by the identification of a suite of endosomal entry factors specific to Sdel virus by a genome-wide CRISPR-Cas9 screen. A panel of host factors regulating the surface expression of ACE2 was identified for both viruses. Using a hamster model, animal-to-animal transmission with the Sdel virus was almost completely abrogated, unlike with Sfull. These findings highlight the critical role of the S1/S2 boundary of the SARS-CoV-2 spike protein in modulating virus entry and transmission.

The outbreak of COVID‐19 coronavirus disease around the end of 2019 has become a pandemic. The preferred method for COVID‐19 detection is the real‐time polymerase chain reaction (RT‐PCR)‐based technique; however, it also has certain limitations, such as sample‐dependent procedures with a relatively high false negative ratio. We propose a safe and efficient method for screening COVID‐19 based on Raman spectroscopy. A total of 177 serum samples are collected from 63 confirmed COVID‐19 patients, 59 suspected cases, and 55 healthy individuals as a control group. Raman spectroscopy is adopted to analyze these samples, and a machine learning support‐vector machine (SVM) method is applied to the spectrum dataset to build a diagnostic algorithm. Furthermore, 20 independent individuals, including 5 asymptomatic COVID‐19 patients and 5 symptomatic COVID‐19 patients, 5 suspected patients, and 5 healthy patients, were sampled for external validation. In these three groups—confirmed COVID‐19, suspected, and healthy individuals—the distribution of statistically significant points of difference showed highly consistency for intergroups after repeated sampling processes. The classification accuracy between the COVID‐19 cases and the suspected cases is 0.87 (95% confidence interval [CI]: 0.85–0.88), and the accuracy between the COVID‐19 and the healthy controls is 0.90 (95% CI: 0.89–0.91), while the accuracy between the suspected cases and the healthy control group is 0.68 (95% CI: 0.67–0.73). For the independent test dataset, we apply the obtained SVM model to the classification of the independent test dataset to have all the results correctly classified. Our model showed that the serum‐level classification results were all correct for independent test dataset. Our results suggest that Raman spectroscopy could be a safe and efficient technique for COVID‐19 screening.

No abstract

In online education systems, for offering proactive services to students (e.g., personalized exercise recommendation), a crucial demand is to predict student performance (e.g., scores) on future exercising activities. Existing prediction methods mainly exploit the historical exercising records of students, where each exercise is usually represented as the manually labeled knowledge concepts, and the richer information contained in the text description of exercises is still underexplored. In this paper, we propose a novel Exercise-Enhanced Recurrent Neural Network (EERNN) framework for student performance prediction by taking full advantage of both student exercising records and the text of each exercise. Specifically, for modeling the student exercising process, we first design a bidirectional LSTM to learn each exercise representation from its text description without any expertise and information loss. Then, we propose a new LSTM architecture to trace student states (i.e., knowledge states) in their sequential exercising process with the combination of exercise representations. For making final predictions, we design two strategies under EERNN, i.e., EERNNM with Markov property and EERNNA with Attention mechanism. Extensive experiments on large-scale real-world data clearly demonstrate the effectiveness of EERNN framework. Moreover, by incorporating the exercise correlations, EERNN can well deal with the cold start problems from both student and exercise perspectives.