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.

Hi-C is commonly used to study three-dimensional genome organization. However, due to the high sequencing cost and technical constraints, the resolution of most Hi-C datasets is coarse, resulting in a loss of information and biological interpretability. Here we develop DeepHiC, a generative adversarial network, to predict high-resolution Hi-C contact maps from low-coverage sequencing data. We demonstrated that DeepHiC is capable of reproducing high-resolution Hi-C data from as few as 1% downsampled reads. Empowered by adversarial training, our method can restore fine-grained details similar to those in high-resolution Hi-C matrices, boosting accuracy in chromatin loops identification and TADs detection, and outperforms the state-of-the-art methods in accuracy of prediction. Finally, application of DeepHiC to Hi-C data on mouse embryonic development can facilitate chromatin loop detection. We develop a web-based tool (DeepHiC, http://sysomics.com/ deephic) that allows researchers to enhance their own Hi-C data with just a few clicks.

Deoxynivalenol (DON), 3-acetyldeoxynivalenol (3-ADON) and 15-acetyldeoxynivalenol (15-ADON) are type B trichothecenes; one of the major pollutants in food and feed products. Although the toxicity of DON has been well documented, information on the toxicity of its acetylated derivative remains incomplete. To acquire more detailed insight into 3-ADON and 15-ADON, Caco-2 cells under 0.5 µM DON, 3-ADON and 15-ADON treatment for 24 h were subjected to RNA-seq analysis. In the present study, 2656, 3132 and 2425 differentially expressed genes (DEGs) were selected, respectively, and were enriched utilizing the Kyoto Encyclopedia of Genes and Genomes (KEGG) and the Gene Ontology (GO) database. The upregulation of ataxia-telangiectasia mutated kinase (ATM), WEE1 homolog 2 (WEE2) and downregulation of proliferating cell nuclear antigen (PCNA), minichromosome maintenance (MCMs), cyclin dependent kinase (CDKs), and E2Fs indicate that the three toxins induced DNA damage, inhibition of DNA replication and cell cycle arrest in Caco-2 cells. Additionally, the upregulation of sestrin (SENEs) and NEIL1 implied that the reason for DNA damage may be attributable to oxidative stress. Our study provides insight into the toxic mechanism of 3-ADON and 15-ADON.

Entity linking refers to the task of aligning mentions of entities in the text to their corresponding entries in a specific knowledge base, which is of great significance for many natural language process applications such as semantic text understanding and knowledge fusion. The pivotal of this problem is how to make effective use of contextual information to disambiguate mentions. Moreover, it has been observed that, in most cases, mention has similar or even identical strings to the entity it refers to. To prevent the model from linking mentions to entities with similar strings rather than the semantically similar ones, in this paper, we introduce the advanced language representation model called BERT (Bidirectional Encoder Representations from Transformers) and design a hard negative samples mining strategy to fine-tune it accordingly. Based on the learned features, we obtain the valid entity through computing the similarity between the textual clues of mentions and the entity candidates in the knowledge base. The proposed hard negative samples mining strategy benefits entity linking from the larger, more expressive pre-trained representations of BERT with limited training time and computing sources. To the best of our knowledge, we are the first to equip entity linking task with the powerful pre-trained general language model by deliberately tackling its potential shortcoming of learning literally, and the experiments on the standard benchmark datasets show that the proposed model yields state-of-the-art results.INDEX TERMS Entity linking, natural language processing (NLP), bidirectional encoder representations from transformers (BERT), deep neural network (DNN).

No abstract