Identification of novel variations in three cases with rare inherited neuromuscular disorder.

Inherited neuromuscular disorder (IND) is a broad-spectrum, clinically diverse group of diseases that are caused due to defects in the neurosystem, muscles and related tissue. Since IND may originate from mutations in hundreds of different genes, the resulting heterogeneity of IND is a great challenge for accurate diagnosis and subsequent management. Three pediatric cases with IND were enrolled in the present study and subjected to a thorough clinical examination. Next, a genetic investigation was conducted using whole-exome sequencing (WES). The suspected variants were validated through Sanger sequencing or quantitative fluorescence PCR assay. A new missense variant of the Spastin (SPAST) gene was found and analyzed at the structural level using molecular dynamics (MD) simulations. All three cases presented with respective specific clinical manifestations, which reflected the diversity of IND. WES detected the diagnostic variants in all 3 cases: A compound variation comprising collagen type VI α3 chain (COL6A3) (NM_004369; exon19):c.6322G>T(p.E1208*) and a one-copy loss of COL6A3:exon19 in Case 1, which are being reported for the first time; a de novo SPAST (NM_014946; exon8):c.1166C>A(p.T389K) variant in Case 2; and a de novo Duchenne muscular dystrophy (NM_004006; exon11):c.1150-17_1160delACTTCCTTCTTTGTCAGGGGTACATGATinsC variant in Case 3. The structural and MD analyses revealed that the detected novel SPAST: c.1166C>A(p.T389K) variant mainly altered the intramolecular hydrogen bonding status and the protein segment's secondary structure. In conclusion, the present study expanded the IND mutation spectrum. The study not only detailed the precise diagnoses of these cases but also furnished substantial grounds for informed consultations. The approach involving the genetic evaluation strategy using WES for variation screening followed by validation using appropriate methods is beneficial due to the considerable heterogeneity of IND.

Heterogeneity analysis of main driving factors affecting potential evapotranspiration changes across different climate regions.

Exploring the influencing factors of potential evapotranspiration (PET) is of great significance for further understanding the causes of climate change and improving agricultural irrigation efficiency. In this study, modified Mann-Kendall analysis was used to elucidate the temporal variation characteristics of meteorological factors and PET based on a dataset from 710 meteorological stations in China. Furthermore, we revealed the main factors that influence the temporal and climate heterogeneity of PET by combining sensitivity analysis with the contribution analysis method. The results showed that 1) climate factors and PET exhibited trend changes on a yearly scale, with slope variation ranges of temperature (T), relative humidity (RH), net radiation (RN), wind speed (U) and PET of 0.03-0.04 °C/a, 0.03-0.08 %/a, 0.001-0.007[MJ/(m2/day)]/a, -0.005 to -0.012(m/s)/a and -0.30-0.38 mm/a, respectively. 2) The sensitivity coefficient fluctuated greatly inter-annually, but the trend was more pronounced inter-annually. Most sensitive factor for PET was RN in hyperarid (HAR), arid (AR) and semiarid regions (SAR), while it changed to RH in semihumid (SHR) and humid regions (HR). PET was more sensitive to RN in dry and relatively wet hot seasons, while it changed to RH during wet and relatively dry cold seasons. 3) PET changes were determined by the relative changes and the sensitivity coefficient, and significant temporal heterogeneity was observed. In HAR, AR, SAR and SHR, the relative changes in T and U result in higher contributions. In HR, PET changes were primarily caused by its higher sensitivity to RH and RN. 4) In dry region and humid-cold seasons, the bigger relative changes of climate factors were the main drivers affecting PET changes, but in humid region and arid-hot seasons, the they were determined by the strong nonlinear relationship between PET and factors. This finding holds great significance for the scientific understanding of the evolution mechanism of PET under changing environments.

Effects of barley seedling powder on rheological properties of dough and quality of steamed bread.

In order to find the optimal share of barley seedling powder (BSP) to improve the rheological properties of wheat dough and physico-chemical properties of steamed bread (SB), BSP was added with wheat flour at various proportions (2-10%). Results showed that with the increasing amount of BSP additive, the farinograph index (86.33-123), dough stability (9.37-12.63 min), and dough development time (6.23-7.63 min) in blend flour increased. Similarly, with the increasing BSP, SB became darker and more greenish, and the total flavonoid content increased. The content of chlorophyll-b, and total chlorophyll demonstrated a faster increase than that of chlorophyll-a. The hardness and chewability of SB improved as well whereas the springiness increased first and then decreased. The best springiness and gumminess of SB were found with 2% and 8% BSP additives respectively. 2%, 4%, and 6% addition of BSP resulted in a slight fluctuation in the bound water quantity than 8% and 10% BSP additive. No new compound formation was confirmed by Infrared analysis and there was only a heat and mass transfer process. Results from this study indicated that SB with improved quality attributes can be prepared from wheat flour fortified with BSP at 2-4%.

STDNet: Rethinking Disentanglement Learning With Information Theory.

Disentangled representation learning is typically achieved by a generative model, variational encoder (VAE). Existing VAE-based methods try to disentangle all the attributes simultaneously in a single hidden space, while the separation of the attribute from irrelevant information varies in complexity. Thus, it should be conducted in different hidden spaces. Therefore, we propose to disentangle the disentanglement itself by assigning the disentanglement of each attribute to different layers. To achieve this, we present a stair disentanglement net (STDNet), a stair-like structure network with each step corresponding to the disentanglement of an attribute. An information separation principle is employed to peel off the irrelevant information to form a compact representation of the targeted attribute within each step. Compact representations, thus, obtained together form the final disentangled representation. To ensure the final disentangled representation is compressed as well as complete with respect to the input data, we propose a variant of the information bottleneck (IB) principle, the stair IB (SIB) principle, to optimize a tradeoff between compression and expressiveness. In particular, for the assignment to the network steps, we define an attribute complexity metric to assign the attributes by the complexity ascending rule (CAR) that dictates a sequencing of the attribute disentanglement in ascending order of complexity. Experimentally, STDNet achieves state-of-the-art results in representation learning and image generation on multiple benchmarks, including Mixed National Institute of Standards and Technology database (MNIST), dSprites, and CelebA. Furthermore, we conduct thorough ablation experiments to show how the strategies employed here contribute to the performance, including neurons block, CAR, hierarchical structure, and variational form of SIB.

A learnable sampling method for scalable graph neural networks.

With the development of graph neural networks, how to handle large-scale graph data has become an increasingly important topic. Currently, most graph neural network models which can be extended to large-scale graphs are based on random sampling methods. However, the sampling process in these models is detached from the forward propagation of neural networks. Moreover, quite a few works design sampling based on statistical estimation methods for graph convolutional networks and the weights of message passing in GCNs nodes are fixed, making these sampling methods not scalable to message passing networks with variable weights, such as graph attention networks. Noting the end-to-end learning capability of neural networks, we propose a learnable sampling method. It solves the problem that random sampling operations cannot calculate gradients and samples nodes with an unfixed probability. In this way, the sampling process is dynamically combined with the forward propagation process of the features, allowing for better training of the networks. And it can be generalized to all message passing models. In addition, we apply the learnable sampling method to GNNs and propose two models. Our method can be flexibly combined with different graph neural network models and achieves excellent accuracy on benchmark datasets with large graphs. Meanwhile, loss function converges to smaller values at a faster rate during training than past methods.

Whole exome sequencing identified a novel compound heterozygous variation in COL7A1 gene causing dystrophic epidermolysis bullosa.

Dystrophic epidermolysis bullosa (DEB) is a series of severe genetic conditions affecting skin and nails caused by mutations in the COL7A1 gene. DEB has a strong phenotypic variability. In the present study, we recruited a case with a boy exhibiting typical DEB indication, and performed a clinical, genetic, and experimental investigation, followed by a prenatal diagnosis on their current pregnancy. Whole exome sequencing identified a novel compound heterozygous variation in COL7A1, consisting of two variants, namely c.191T>C (p.Leu64Pro) and c.5124G>A (p.Leu1708=) in the proband. In vitro study by minigene system indicated that c.5124G>A would result in an increased ratio of a transcript with exon-skipping, which supported its pathogenicity. Further prenatal detection confirmed the genotype-phenotye co-separation in this family. In conclusion, the findings in our study expanded the mutation spectrum of DEB, and emphasized the importance of paying attention to specific synonymous variants in the filtering process.

An accelerated proximal gradient algorithm for singly linearly constrained quadratic programs with box constraints.

Recently, the existed proximal gradient algorithms had been used to solve non-smooth convex optimization problems. As a special nonsmooth convex problem, the singly linearly constrained quadratic programs with box constraints appear in a wide range of applications. Hence, we propose an accelerated proximal gradient algorithm for singly linearly constrained quadratic programs with box constraints. At each iteration, the subproblem whose Hessian matrix is diagonal and positive definite is an easy model which can be solved efficiently via searching a root of a piecewise linear function. It is proved that the new algorithm can terminate at an ε-optimal solution within [Formula: see text] iterations. Moreover, no line search is needed in this algorithm, and the global convergence can be proved under mild conditions. Numerical results are reported for solving quadratic programs arising from the training of support vector machines, which show that the new algorithm is efficient.