Image sequence-based risk behavior detection of power operation inspection personnel

A novel image sequence-based risk behavior detection method to achieve high-precision risk behavior detection for power maintenance personnel is proposed in this paper. In this method, the original image sequence data is first separated from the foreground and background. Then, the free anchor frame detection method is used in the foreground image to detect the personnel and correct their direction. Finally, human posture nodes are extracted from each frame of the image sequence, which are then used to identify the abnormal behavior of the human. Simulation experiment results demonstrate that the proposed algorithm has significant advantages in terms of the accuracy of human posture node detection and risk behavior identification.     

Removing data and using metafounders alleviates biases for all traits in Lacaune dairy sheep predictions

Bias in dairy genetic evaluations, when it exists, has to be understood and properly addressed. The origin of biases is not always clear. We analyzed 40 yr of records from the Lacaune dairy sheep breeding program to evaluate the extent of bias, assess possible corrections, and emit hypotheses on its origin. The data set included 7 traits (milk yield, fat and protein contents, somatic cell score, teat angle, udder cleft, and udder depth) with records from 600,000 to 5 million depending on the trait, ～1,900,000 animals, and ～5,900 genotyped elite artificial insemination rams. For the ～8% animals with missing sire, we fit 25 unknown parent groups. We used the linear regression method to compare “partial” and “whole” predictions of young rams before and after progeny testing, with 7 cut-off points, and we obtained estimates of their bias, (over)dispersion, and accuracy in early proofs. We tried (1) several scenarios as follows: multiple or single trait, the “official” (routine) evaluation, which is a mixture of both single and multiple trait, and “deletion” of data before 1990; and (2) several models as follows: BLUP and single-step genomic (SSG)BLUP with fixed unknown parent groups or metafounders, where, for metafounders, their relationship matrix gamma was estimated using either a model for inbreeding trend, or base allele frequencies estimated by peeling. The estimate of gamma obtained by modeling the inbreeding trend resulted in an estimated increase of inbreeding, based on markers, faster than the pedigree-based one. The estimated genetic trends were similar for most models and scenarios across all traits, but were shrunken when gamma was estimated by peeling. This was due to shrinking of the estimates of metafounders in the latter case. Across scenarios, all traits showed bias, generally as an overestimate of genetic trend for milk yield and an underestimate for the other traits. As for the slope, it showed overdispersion of estimated breeding values for all traits. Using multiple-trait models slightly reduced the overestimate of genetic trend and the overdispersion, as did including genomic information (i.e., SSGBLUP) when the gamma matrix was estimated by the model for inbreeding trend. However, only deletion of historical data before 1990 resulted in elimination of both kind of biases. The SSGBLUP resulted in more accurate early proofs than BLUP for all traits. We considered that a snowball effect of small errors in each genetic evaluation, combined with selection, may have resulted in biased evaluations. Improving statistical methods reduced some bias but not all, and a simple solution for this data set was to remove historical records.     

Cloning,sequencing and structural analysis of membrane-bound polyphenol oxidase from Granny Smith apples (Malus × domestica Borkh)

The gene sequence coding for the membrane-bound polyphenol oxidase (mPPO) with a length of 1761 bp was cloned by PCR method and shown to contain one highly conserved sequence encoding a di-copper-binding region. The predicted three-dimensional structure of mPPO indicated that the active site was located near two copper ions and composed of a typical bundle of four α-helices. Each of the two catalytic copper ions was coordinated with three histidine residues in the hydrophobic pocket, yielding His 180, His 201, His 210, His 332, His 336 and His 366. Docking studies showed that 4-methylcatechol and chlorogenic acid have different binding models due to different ligand sizes and binding sites in the active centre, and it was found that the smaller compound exhibited a higher affinity for mPPO. Molecular dynamic simulation results indicated that Phe 353 is important in controlling enzymatic activity through influencing substrate coordination in the active site.     

Bacteriophages vB_Sen-TO17 and vB_Sen-E22, Newly Isolated Viruses from Chicken Feces,Specific for Several Salmonella enterica Strains

Two newly discovered bacteriophages, isolated from chicken feces and infecting Salmonella enterica strains, are described in this report. These phages have been named vB_Sen-TO17 and vB_Sen-E22, and we present their molecular and functional characterization. Both studied viruses are able to infect several S. enterica strains and develop lytically, but their specific host ranges differ significantly. Electron microscopic analyses of virions have been performed, and full genome sequences were determined and characterized, along with molecular phylogenetic studies. Genomes of vB_Sen-TO17 (ds DNA of 41,658 bp) and vB_Sen-E22 (dsDNA of 108,987 bp) are devoid of homologs of any known or putative gene coding for toxins or any other proteins potentially deleterious for eukaryotic cells. Both phages adsorbed efficiently (>95% adsorbed virions) within 10 min at 42 °C (resembling chicken body temperature) on cells of most tested host strains. Kinetics of lytic development of vB_Sen-TO17 and vB_Sen-E22, determined in one-step growth experiments, indicated that development is complete within 30–40 min at 42 °C, whereas burst sizes vary from 9 to 79 progeny phages per cell for vB_Sen-TO17 and from 18 to 64 for vB_Sen-E22, depending on the host strain. Virions of both phages were relatively stable (from several percent to almost 100% survivability) under various conditions, including acidic and alkaline pH values (from 3 to 12), temperatures from −80 °C to 60 °C, 70% ethanol, chloroform, and 10% DMSO. These characteristics of vB_Sen-TO17 and vB_Sen-E22 indicate that these phages might be considered in further studies on phage therapy, particularly in attempts to eliminate S. enterica from chicken intestine.     

Compilation of Modern Technologies To Map Genome-Wide Cytosine Modifications in DNA

Over the past few decades, various DNA modification detection methods have been developed; many of the high-resolution methods are based on bisulfite treatment, which leads to DNA degradation, to a degree. Thus, novel bisulfite-free approaches have been developed in recent years and shown to be useful for epigenome analysis in otherwise difficult-to-handle, but important, DNA samples, such as hmC-seal and hmC-CATCH. Herein, an overview of advances in the development of epigenome sequencing methods for these important DNA modifications is provided.     

Effect of ZnO/K2O ratio on the crystallization sequence and microstructure of lithium disilicate glass-ceramics

The effect of ZnO/K₂O (Z/K) ratio on the crystallization sequence and microstructure of lithium disilicate (Li₂Si₂O₅: LS2) glass-ceramics was carefully investigated for the SiO₂-Li₂O-K₂O-ZnO-P₂O₅ system. The Z/K ratios of precursor glasses were varied from 0 to 3.5 while the nucleating agent of P₂O₅ and glass modifiers of ZnO plus K₂O were fixed to have 1.5 and 4.5 mol% relative to LS2, respectively. For the samples prepared by two-stage heat treatments of 500 °C for 1 h and 800 °C for 2 h in air, the LS2 nucleation rate was increased with increasing the Z/K ratio due to the variation in crystallization sequence from type II (Li₂SiO₃: LS) to type I (LS + LS2) in addition to an amorphous phase separation in base glass. Consequently, with increasing the Z/K ratio, the LS2 crystalline phase within the glass matrix continuously changed from larger acicular ones to smaller equiaxed ones.     

Relatedness between numerically small Dutch Red dairy cattle populations and possibilities for multibreed genomic prediction

Red dairy breeds are a valuable cultural and historical asset, and often a source of unique genetic diversity. However, they have difficulties competing with other, more productive, dairy breeds. Improving competitiveness of Red dairy breeds, by accelerating their genetic improvement using genomic selection, may be a promising strategy to secure their long-term future. For many Red dairy breeds, establishing a sufficiently large breed-specific reference population for genomic prediction is often not possible, but may be overcome by adding individuals from another breed. Relatedness between breeds strongly decides the benefit of adding another breed to the reference population. To prioritize among available breeds, the effective number of chromosome segments (M_e) can be used as an indicator of relatedness between individuals from different breeds. The M_e is also an important parameter in determining the accuracy of genomic prediction. The M_e can be estimated both within a population and between 2 populations or breeds, as the reciprocal of the variance of genomic relationships. We investigated relatedness between 6 Dutch Red cattle breeds, Groningen White Headed (GWH), Dutch Friesian (DF), Meuse-Rhine-Yssel (MRY), Dutch Belted (DB), Deep Red (DR), and Improved Red (IR), focusing primarily on the M_e, to predict which of those breeds may benefit from including reference animals of the other breeds. All of these breeds, except MRY, are under high risk of extinction. Our results indicated high variability of M_e, especially between M_e ranging from ～3,500 to ～17,400, indicating different levels of relatedness between the breeds. Two clusters are especially important, one formed by MRY, DR, and IR, and the other comprising DF and DB. Although relatedness between breeds within each of these 2 clusters is high, across-breed genomic prediction is still limited by the current number of genotyped individuals, which for many breeds is low. However, adding MRY individuals would increase the reference population of DR substantially. We estimated that between 11 and 133 individuals from other breeds are needed to achieve accuracy of genomic prediction equivalent to using one additional individual from the same breed. Given the variation in size of the breeds in this study, the benefit of a multibreed reference population is expected to be lower for larger breeds than for the smaller ones.     

Genomic evaluation of dairy heifer livability

Differences in breeds and sire lines suggest the presence of a genetic component for heifer livability (HLIV). Genomic evaluation for this trait can increase profitability and improve animal health and welfare. Evaluations for HLIV were examined from 3,362,499 calf data records from heifers of all breeds born from 2009 to 2016. Data were obtained from the national cooperator database maintained by the Council on Dairy Cattle Breeding (https://www.uscdcb.com/). The total number of deaths reported was 134,753 (4.01%), which included herds with death loss between 1.5 and 25.5%. Age at death was evaluated and ranged from >2 d of age until the heifer left the herd, with a maximum of 18 mo of age. Records were not included until 3 yr after the birthdate so that live status of contemporaries could be confirmed by a calving date for those animals. Deaths observed until 2 d after birth were considered to be a stillbirth rather than a failure of HLIV. The scale used for analysis of HLIV was 0 (died) or 100 (live), and the heritability estimate was 0.7% based on sire model with restricted maximum likelihood estimation. Genomic predicted transmitting abilities for Holstein ranged from ?1.6% to +1.6% with a standard deviation of 0.5%, and genomic predicted transmitting abilities for Jersey ranged from ?0.5% to +0.5% with a standard deviation of 0.2%. The mean overall death loss was about 4%. Reliabilities of genomic predictions for young animals averaged 46% for Holsteins and 30% for Jerseys, and corresponding traditional parent average reliabilities averaged 16% and 12%, respectively. Correlations of HLIV were 0.44 with productive life, 0.18 to 0.22 with yield traits, and 0.29 with early first calving on proven Holstein bulls. The HLIV trait had a favorable genetic trend in recent years, likely because of the indirect selection associated with the correlated traits. The trait HLIV should receive 1% of emphasis on the Lifetime Net Merit index, resulting in economic progress worth $50,000/yr. By encouraging more comprehensive recording on calf mortality, the reliabilities of genetic predictions could increase significantly.     

基于深度学习的文本自动摘要方案

针对自然语言处理（NLP）生成式自动摘要领域的语义理解不充分、摘要语句不通顺和摘要准确度不够高的问题，提出了一种新的生成式自动摘要解决方案，包括一种改进的词向量生成技术和一个生成式自动摘要模型。改进的词向量生成技术以Skip-Gram方法生成的词向量为基础，结合摘要的特点，引入词性、词频和逆文本频率三个词特征，有效地提高了词语的理解；而提出的Bi-MulRnn+生成式自动摘要模型以序列映射（seq2seq）与自编码器结构为基础，引入注意力机制、门控循环单元（GRU）结构、双向循环神经网络（BiRnn）、多层循环神经网络（MultiRnn）和集束搜索，提高了生成式摘要准确性与语句流畅度。基于大规模中文短文本摘要（LCSTS）数据集的实验结果表明，该方案能够有效地解决短文本生成式摘要问题，并在Rouge标准评价体系中表现良好，提高了摘要准确性与语句流畅度。