CaFANet: Causal-Factors-Aware Attention Networks for Equipment Fault Prediction in the Internet of Things.

Existing fault prediction algorithms based on deep learning have achieved good prediction performance. These algorithms treat all features fairly and assume that the progression of the equipment faults is stationary throughout the entire lifecycle. In fact, each feature has a different contribution to the accuracy of fault prediction, and the progress of equipment faults is non-stationary. More specifically, capturing the time point at which a fault first appears is more important for improving the accuracy of fault prediction. Moreover, the progress of the different faults of equipment varies significantly. Therefore, taking feature differences and time information into consideration, we propose a Causal-Factors-Aware Attention Network, CaFANet, for equipment fault prediction in the Internet of Things. Experimental results and performance analysis confirm the superiority of the proposed algorithm over traditional machine learning methods with prediction accuracy improved by up to 15.3%.

Genome-Wide Characterization of HSP90 Gene Family in Cucumber and Their Potential Roles in Response to Abiotic and Biotic Stresses.

Heat shock protein 90 (HSP90) possesses critical functions in plant developmental control and defense reactions. The HSP90 gene family has been studied in various plant species. However, the HSP90 gene family in cucumber has not been characterized in detail. In this study, a total of six HSP90 genes were identified from the cucumber genome, which were distributed to five chromosomes. Phylogenetic analysis divided the cucumber HSP90 genes into two groups. The structural characteristics of cucumber HSP90 members in the same group were similar but varied among different groups. Synteny analysis showed that only one cucumber HSP90 gene, Csa1G569290, was conservative, which was not collinear with any HSP90 gene in Arabidopsis and rice. The other five cucumber HSP90 genes were collinear with five Arabidopsis HSP90 genes and six rice HSP90 genes. Only one pair of paralogous genes in the cucumber HSP90 gene family, namely one pair of tandem duplication genes (Csa1G569270/Csa1G569290), was detected. The promoter analysis showed that the promoters of cucumber HSP90 genes contained hormone, stress, and development-related cis-elements. Tissue-specific expression analysis revealed that only one cucumber HSP90 gene Csa3G183950 was highly expressed in tendril but low or not expressed in other tissues, while the other five HSP90 genes were expressed in all tissues. Furthermore, the expression levels of cucumber HSP90 genes were differentially induced by temperature and photoperiod, gibberellin (GA), downy mildew, and powdery mildew stimuli. Two cucumber HSP90 genes, Csa1G569270 and Csa1G569290, were both differentially expressed in response to abiotic and biotic stresses, which means that these two HSP90 genes play important roles in the process of cucumber growth and development. These findings improve our understanding of cucumber HSP90 family genes and provide preliminary information for further studies of cucumber HSP90 gene functions in plant growth and development.

Divergent gross nitrogen transformation paths in the topsoil and subsoil between abandoned and agricultural cultivation land in irrigated areas.

The nitrate concentration in groundwater has increased in many irrigated areas worldwide due to the excessive use of both water and fertilizers. Abandoned farmlands in such irrigated areas may alter the nitrogen (N) cycle because of drastically changed water and N inputs. However, the mechanisms of the N cycle in response to such changes remain unclear. We studied biogeochemical N cycling and microbiological responses from abandoned arable lands (AF), for the topsoil (20 cm depth) and subsoil (100 cm depth) layers, in comparison with irrigation-fertilization (control = CK) land, by using 15N tracing techniques, the 16S rRNA gene, and real-time PCR (qPCR) to reveal the mechanisms underpinning the N cycle. We found that the biogeochemical environment of abandoned soils shifted their N-cycling pathways. Except for reduced soil moisture, soil properties of total C and N, as well pH, showed improvement in the two layers of AF. But the microbial abundances of ammonia-oxidizing bacteria (AOB-amoA), archaea (AOA-amoA), bacteria and fungi were all significantly lower in the AF; and they presented a consistent trend in the subsoil of the two lands. Significant differences in gross N transformation rates were found for mineralization rates (MN) and autotrophic nitrification rate (ONH4) between lands or depths. Compared with AF, MN was increased by 1.45- and 11.75-times, and ONH4 by 1.69- and 2.89-times in the topsoil and subsoil of CK, respectively. Our results suggest that the SM × C/N interaction provides insight into the mechanisms underlying the soil microbe-driven changes to transformation rates in nitrogen dynamics after abandoning water-limited lands. The high moisture and N inputs reported here highlight the dynamics and prevalence of MN and ONH4, and an increasing the nitrate leaching rate in the unsaturated zone, which poses a major threat to groundwater quality.