To determine molecular viral components which can induce innate immune responses in human peripheral blood mononuclear cells (PBMC), we investigated the anti-neoplastic agent Newcastle disease virus (NDV) and its two spike proteins hemagglutinin-neuraminidase (HN) and fusion protein (F). NDV was an excellent inducer in PBMC of IFN-alpha production and capable of inducing upregulation of plasma membrane expression of tumor necrosis factor related apoptosis inducing ligand (TRAIL). Viral replication was not required for these responses because NDV inactivated for 5 min by UV was as good as live NDV. NDV-modified and paraformaldehyde-fixed BHK cells could also trigger IFN-alpha and TRAIL induction, indicating that contacts of responder cells with NDV-modified cell surfaces are sufficient to induce these activities in PBMC. Antibodies against HN but not F were able to block these responses. Finally we could show that HN but not F induced IFN-alpha and TRAIL in PBMC. This was possible through the use of respective gene transfectants generated with the help of Semliki Forest virus (SFV) replicase-based DNA recombinant expression systems. Upon contact with BHK cells expressing HN but not F at their cell surface, human PBMC produced IFN-alpha and some cells, including monocytes and T lymphocytes, upregulated cell surface TRAIL expression.

With great practical value, the study of Multidomain Neural Machine Translation (NMT) mainly focuses on using mixed-domain parallel sentences to construct a unified model that allows translation to switch between different domains. Intuitively, words in a sentence are related to its domain to varying degrees, so that they will exert disparate impacts on the multi-domain NMT modeling. Based on this intuition, in this paper, we devote to distinguishing and exploiting word-level domain contexts for multi-domain NMT. To this end, we jointly model NMT with monolingual attention-based domain classification tasks and improve NMT as follows: 1) Based on the sentence representations produced by a domain classifier and an adversarial domain classifier, we generate two gating vectors and use them to construct domain-specific and domain-shared annotations, for later translation predictions via different attention models; 2) We utilize the attention weights derived from target-side domain classifier to adjust the weights of target words in the training objective, enabling domain-related words to have greater impacts during model training. Experimental results on Chinese-English and English-French multi-domain translation tasks demonstrate the effectiveness of the proposed model. Source codes of this paper are available on Github https://github.com/DeepLearnXMU/WDCNMT.

Abstract. The monthly standardized precipitation evapotranspiration index (SPEI) can be used to monitor and assess drought characteristics with 1-month or longer drought duration. Based on data from 1961 to 2018 at 427 meteorological stations across mainland China, we developed a daily SPEI dataset to overcome the shortcoming of the coarse temporal scale of monthly SPEI. Our dataset not only can be used to identify the start and end dates of drought events, but also can be used to investigate the meteorological, agricultural, hydrological, and socioeconomic droughts with a different timescales. In the present study, the SPEI data with 3-month (about 90 d) timescale were taken as a demonstration example to analyze spatial distribution and temporal changes in drought conditions for mainland China. The SPEI data with a 3-month (about 90 d) timescale showed no obvious intensifying trends in terms of severity, duration, and frequency of drought events from 1961 to 2018. Our drought dataset serves as a unique resource with daily resolution to a variety of research communities including meteorology, geography, and natural hazard studies. The daily SPEI dataset developed is free, open, and publicly available from this study. The dataset with daily SPEI is publicly available via the figshare portal (Wang et al., 2020c), with https://doi.org/10.6084/m9.figshare.12568280.Highlights. A multi-scale daily SPEI dataset was developed across mainland China from 1961 to 2018. The daily SPEI dataset can be used to identify the start and end days of the drought event. The developed daily SPEI dataset in this study is free, open, and publicly available.

Sentence ordering is to restore the original paragraph from a set of sentences. It involves capturing global dependencies among sentences regardless of their input order. In this paper, we propose a novel and flexible graph-based neural sentence ordering model, which adopts graph recurrent network to accurately learn semantic representations of the sentences.Instead of assuming connections between all pairs of input sentences, we use entities that are shared among multiple sentences to make more expressive graph representations with less noise. Experimental results show that our proposed model outperforms the existing stateof-the-art systems on several benchmark datasets, demonstrating the effectiveness of our model. We also conduct a thorough analysis on how entities help the performance. Our code is available at