Here, we developed a novel deep network architecture called the multi-layered Long-Short Term Memory networks (LSTMs) approach for the prediction of protein interface residue pairs. Firstly, we created three new descriptions and used other six worked characterizations to describe an amino acid, then we employed these features to discriminate between interface residue pairs and non-interface residue pairs. Secondly, we used two thresholds to select residue pairs that are more likely to be interface residue pairs. Furthermore, this step increases the proportion of interface residue pairs and reduces the influence of imbalanced data. Thirdly, we built deep network architectures based on Long-Short Term Memory networks algorithm to organize and refine the prediction of interface residue pairs by employing features mentioned above. We trained the deep networks on dimers in the unbound state in the international Protein-protein Docking Benchmark version 3.0. The updated data sets in the versions 4.0 and 5.0 were used as the validation set and test set respectively. For our best model, the accuracy rate was over 62% when we chose the top 0.2% pairs of every dimer in the test set as predictions, which will be very helpful for the understanding of protein-protein interaction mechanisms and for guidance in biological experiments.

DNA phosphorothioate (PT) modification, with a nonbridging phosphate oxygen substituted by sulfur, represents a widespread epigenetic marker in prokaryotes and provides protection against genetic parasites. In the PT-based defense system Ssp, SspABCD confers a single-stranded PT modification of host DNA in the 5′-CPSCA-3′ motif and SspE impedes phage propagation. SspE relies on PT modification in host DNA to exert antiphage activity. Here, structural and biochemical analyses reveal that SspE is preferentially recruited to PT sites mediated by the joint action of its N-terminal domain (NTD) hydrophobic cavity and C-terminal domain (CTD) DNA binding region. PT recognition enlarges the GTP-binding pocket, thereby increasing GTP hydrolysis activity, which subsequently triggers a conformational switch of SspE from a closed to an open state. The closed-to-open transition promotes the dissociation of SspE from self PT-DNA and turns on the DNA nicking nuclease activity of CTD, enabling SspE to accomplish self-nonself discrimination and limit phage predation, even when only a small fraction of modifiable consensus sequences is PT-protected in a bacterial genome.

Air quality issue such as particulate matter pollution (PM 2.5 and PM 10) has become one of the biggest environmental problem in China. As one of the most important industrial base and economic core regions of China, Northeast China is facing serious air pollution problems in recent years, which has a profound impact on the health of local residents and atmospheric environment in some part of East Asia. Therefore, it is urgent to understand temporal-spatial characteristics of particles and analyze the causality factors. The results demonstrated that variation trend of particles was almost similar, the annual, monthly and daily distribution had their own characteristics. Particles decreased gradually from south to north, from west to east. Correlation analysis showed that wind speed was the most important factor affecting particles, and temperature, air pressure and relative humidity were key factors in some seasons. Path analysis showed that there was complex unidirectional causal relationship between particles and individual or combined effects, and NO 2 and CO were key factors affecting PM 2.5. The hot and cold areas changed little with the seasons. All the above results suggests that planning the industrial layout, adjusting industrial structure, joint prevention and control were necessary measure to reduce particles concentration. Air pollution has become one of the severe environmental problems in China. People in China have to cope with high levels of PM 2.5 (Particulate matter smaller than 2.5 micrometers in diameter). It not only induces the increase of low-visibility days 1-3 , but also penetrates lungs and does harm to respiration, cardiovascular, cerebral vascular and nervous systems 4-7. Studies show that atmospheric particulate matter is closely related to mortality for lots of causes 8-11. In addition, Particulate matter possibly has large influence on regional and global climate change 12-14 , ecosystems 15 , economic development 16 and so on. Thus, Chinese government has established air quality monitoring stations in many cities to monitor pollutants (including PM 2.5 , PM 10 , SO 2 , NO 2 , CO and O 3) mass concentration, published the relative observed data online from January 2013. These air quality data played an important role on analyzing local air pollution situation and establishing air pollutant mass concentration prediction model 17. As previous studies have shown, pollutants sources emission, external transport, meteorological conditions and secondary generation of air pollution were important factors influencing atmospheric particulate matter mass concentration. In particular, meteorological conditions can diffuse, dilute and accumulate air pollutant mass concentration at a large extent. Secondary generation will aggravate air pollutants mass concentration 18,19. Therefore, the study of the distribution characteristics of air pollutants, the relationship between meteorological conditions and air pollutants mass concentration, as well as the relationship between different air pollutants...