Search

Although persistent efforts have identified and characterized a few candidate genes and related biological processes with potential functions in the adaptation of many species to extreme environments, few works have been conducted to determine the genomic basis of adaptation in endangered livestock breeds that have been living in extreme conditions for more than thousands of years. To fill this gap, we sequenced the whole genomes of nine individuals from three Chinese native endangered cattle breeds that are living in high‐altitude or arid environments. Phylogenetic and evolutionary history analyses of these three and other six breeds showed that the genetic structure of the cattle populations is primarily related to geographic location. Interestingly, we identified pervasive introgression from the yak to Zhangmu cattle (ZMC) that cover several genes (e.g., NOS2, EGLN1 and EPAS1) involved in the hypoxia response and previously identified as positive selection genes in other species, which suggested that the adaptive introgression from yak may have contributed to the adaptation of ZMC to high‐altitude environments. In addition, by contrasting the breeds in opposite living conditions, we revealed a set of candidate genes with various functions from hypoxia response, water metabolism, immune response and body shape change to embryo development and skeletal system development, etc., that may be related to high‐altitude or arid adaptation. Our research provides new insights into the recovery and adaptation of endangered native cattle and other species in extreme environments and valuable resources for future research on animal husbandry to cope with climate change.

Though recent methods on semi-supervised video object segmentation (VOS) have achieved an appreciable improvement of segmentation accuracy, it is still hard to get an adequate speed-accuracy balance when facing real-world application scenarios. In this work, we propose Discriminative Matching for real-time Video Object Segmentation (DMVOS), a real-time VOS framework with high-accuracy to fill this gap. Based on the matching mechanism, our framework introduces discriminative information through the Isometric Correlation module and the Instance Center Offset module. Specifically, the isometric correlation module learns a pixel-level similarity map with semantic discriminability, and the instance center offset module is applied to exploit the instance-level spatial discriminability. Experiments on two benchmark datasets show that our model achieves state-of-the-art performance with extremely fast speed, for example, J&F of 87.8% on DAVIS-2016 validation set with 35 milliseconds per frame.

Distributed wind power (DWP) needs to be consumed locally under a 110 kV network without reverse power flow in China. To maximize the use of DWP, this paper proposes a novel method for capacity planning of DWP with participation of the energy storage system (ESS) in multiple scenarios by means of a variable-structure copula and optimization theory. First, wind power and local load are predicted at the planning stage by an autoregressive moving average (ARMA) model, then, variable-structure copula models are established based on different time segment strategies to depict the correlation of DWP and load, and the joint typical scenarios of DWP and load are generated by clustering, and a capacity planning model of DWP is proposed considering investment and operation cost, and environmental benefit and line loss cost under typical scenario conditions. Moreover, a collaborative capacity planning model for DWP and ESS is prospectively proposed. Based on the modified IEEE-33 bus system, the results of the case study show that the DWP capacity result is more reasonable after considering the correlation of wind and load by using a variable-structure copula. With consideration of the collaborative planning of DWP and load, the consumption of DWP is further improved, the annual cost of the system is more economical, and the quality of voltage is effectively improved. The study results validate the proposed method and provide effective reference for the planning strategy of DWP.

The scale of training data is significant in segmentation task, especially in segmenting the medical coronary artery angiograms. Traditional semantic segmentation networks have been restricted in this field, due to the particularity of cardiac coronary angiography data, that is, it is very difficult to balance the manual labeling costs and network accuracy. On the basis of these observations, we propose a new method to generate the so-called ‘pseudo-precise’ label and a complementary training pipeline, which can improve the performance of the networks on the premise of reducing labor costs as much as possible. Our method can thus increase the f1-score by 4%-11% with the same amount of precisely labeled data.

Network biology is conceptualized as an interdisciplinary field, lying at the intersection among graph theory, statistical mechanics and biology. Great efforts have been made to promote the concept of network biology and its various applications in life s