Rapid Mining of Fast Ion Conductors via Subgraph Isomorphism Matching.

The rapidly evolving field of inorganic solid-state electrolytes (ISSEs) has been driven in recent years by advances in data-mining techniques, which facilitates the high-throughput computational screening for candidate materials in the databases. The key to the mining process is the selection of critical features that underline the similarity of a material to an existing ISSE. Unfortunately, this selection is generally subjective and frequently under debate. Here we propose a subgraph isomorphism matching method that allows an objective evaluation of the similarity between two compounds according to the topology of the local atomic environment. The matching algorithm has been applied to discover four structure types that are highly analogous to the LiTi2(PO4)3 NASICON prototype. We demonstrate that the local atomic environments similar to LiTi2(PO4)3 endow these four structures with favorable Li diffusion tunnels and ionic conductivity on par with those of the prototype. By further taking into account the electronic structure and electrochemical stability window, 13 compounds are identified to be potential ISSEs. Our findings not only offer a promising approach toward rapid mining of fast ion conductors without limitation in the compositional range but also reveal insights into the design of ISSEs according to the topology of their framework structures.

Crystal Structure Assignment for Unknown Compounds from X-ray Diffraction Patterns with Deep Learning.

Determining the structures of previously unseen compounds from experimental characterizations is a crucial part of materials science. It requires a step of searching for the structure type that conforms to the lattice of the unknown compound, which enables the pattern matching process for characterization data, such as X-ray diffraction (XRD) patterns. However, this procedure typically places a high demand on domain expertise, thus creating an obstacle for computer-driven automation. Here, we address this challenge by leveraging a deep-learning model composed of a union of convolutional residual neural networks. The accuracy of the model is demonstrated on a dataset of over 60,000 different compounds for 100 structure types, and additional categories can be integrated without the need to retrain the existing networks. We also unravel the operation of the deep-learning black box and highlight the way in which the resemblance between the unknown compound and a structure type is quantified based on both local and global characteristics in XRD patterns. This computational tool opens new avenues for automating structure analysis on materials unearthed in high-throughput experimentation.

Comparative study on the structural and in vitro digestion properties of starch within potato parenchyma cells under different cooking methods.

To study the effects of cooking methods on the structure and digestion changes of starch encapsulated by cellular structure, intact potato parenchyma cells were successfully isolated and then subjected to different domestic cooking methods, including baking, frying, boiling, and autoclaving. The morphology, crystalline structure, thermal properties, and in vitro starch digestibility of cooked cell samples were investigated. Our results indicated that potato cell walls remained intact and performed as physical barriers preventing the diffusion/absorption of α-amylase to intracellular starch substrates after baking or frying treatment. However, boiling or autoclaving treatment destroyed cell wall structure, and the disrupted cellular structure reduced the digestion rate, likely by inhibiting diffusion of amylase through a weakened cell wall barrier, but could not lower the final digestion extent when compared to the pure starch. These findings suggested that potato products with lower glycemic index can be obtained by baking or frying treatment.

Multi-class GAN for generating multi-class images in object recognition.

The current generative adversarial network (GAN) is limited in the application of data augmentation in object recognition. The training of the GAN is unstable, and the generated image quality is poor. Methods such as progressive growing of GANs and multi-scale gradient GAN solve these problems. The packed GAN (PacGAN) solves the problem of mode collapse during training. However, these methods can generate only one type of image at a time, and the training time is long. To solve the above problems, this paper proposes the multi-class GAN (Mc-GAN). It uses an augmented discriminator to train multiple generators at the same time. Through iterative training, the discriminator can accurately judge the output of each generator and guide it to generate the corresponding image. This paper analyzes the optimization process of the objective function of Mc-GAN. Experiments show that the method can generate high-quality images and reduce training time, and it can be used for data augmentation in object recognition. It effectively improves the practicality of GAN.

Transcriptomic responses of the clam Meretrix meretrix to the organophosphorus pesticide (dimethoate).

Organophosphorus pesticides (OPs) play a certain role in promoting the development of agriculture and forestry, but they may cause potential harm to aquatic life when entering rivers and polluting water sources. Previous researches have shown that OPs participate in the regulation mechanism of aquatic organisms. Here, our aim is to determine the underlying mechanisms of one OP (dimethoate) at the transcriptional level using the clam Meretrix meretrix. 4119 DEGs were obtained from high-throughput RNA sequencing data. Then, expression profiles of some genes were verified by qPCR, which showed a positive correlation with the RNA sequencing results. 14,481 simple sequence repeats were also identified and could be further used as molecular markers. In addition, some oxidative, immune, and stress-related genes were further discussed and could also be used as biomarkers to indicate the biological response of dimethoate. This study will help to better understand the clam's response mechanism to dimethoate stress.

BnaMPK3 Is a Key Regulator of Defense Responses to the Devastating Plant Pathogen Sclerotinia sclerotiorum in Oilseed Rape.

The disease caused by Sclerotinia sclerotiorum has traditionally been difficult to control, resulting in tremendous economic losses in oilseed rape (Brassica napus). Identification of important genes in the defense responses is critical for molecular breeding, an important strategy for controlling the disease. Here, we report that a B. napus mitogen-activated protein kinase gene, BnaMPK3, plays an important role in the defense against S. sclerotiorum in oilseed rape. BnaMPK3 is highly expressed in the stems, flowers and leaves, and its product is localized in the nucleus. Furthermore, BnaMPK3 is highly responsive to infection by S. sclerotiorum and treatment with jasmonic acid (JA) or the biosynthesis precursor of ethylene (ET), but not to treatment with salicylic acid (SA) or abscisic acid. Moreover, overexpression (OE) of BnaMPK3 in B. napus and Nicotiana benthamiana results in significantly enhanced resistance to S. sclerotiorum, whereas resistance is diminished in RNAi transgenic plants. After S. sclerotiorum infection, defense responses associated with ET, JA, and SA signaling are intensified in the BnaMPK3-OE plants but weakened in the BnaMPK3-RNAi plants when compared to those in the wild type plants; by contrast the level of both H2O2 accumulation and cell death exhibits a reverse pattern. The candidate gene association analyses show that the BnaMPK3-encoding BnaA06g18440D locus is a cause of variation in the resistance to S. sclerotiorum in natural B. napus population. These results suggest that BnaMPK3 is a key regulator of multiple defense responses to S. sclerotiorum, which may guide the resistance improvement of oilseed rape and related economic crops.

Immunohistochemical profile and prognostic significance in primary central nervous system lymphoma: Analysis of 89 cases.

The majority of primary central nervous system lymphomas (PCNSLs) are diffuse large B cell lymphoma, characterized by poor prognosis. In the present study, the expression of cluster of differentiation (CD)10, B cell lymphoma (BCL)-6, multiple myeloma-1 (MUM-1), BCL-2, CD138 and Ki-67 was analyzed by immunohistochemistry in 89 Chinese PCNSL cases, and the potential prognostic significance was evaluated. CD10, BCL-6, MUM-1, BCL-2 and CD138 were positive in 16.9 (15/89), 51.7 (46/89), 92.1 (82/89), 73.3 (63/86) and 0% (0/65) of all cases, respectively. According to the Hans algorithm, 71 patients (79.8%) were classified into the non-germinal center B cell-like (non-GCB) group, indicating a post-germinal center origin of PCNSL. The median follow-up time of 73 patients was 13 months [95% confidence interval (CI), 10.93-15.08]. The median overall survival (OS) time was 45.3 months (95% CI, 25.01-65.59) and the median progression-free survival (PFS) time was 30.0 months (95% CI, 13.43-46.57). Age (>60 years) was associated with a shorter OS time (P=0.009). Ki-67 (cutoff point 90%) was associated with shorter OS (P=0.037) and shorter PFS (P=0.039) times. No other immunohistochemical markers were associated with prognosis. On multivariate analysis, age (>60 years) was associated with shorter OS time (P=0.038), but immunophenotype and expression status of Ki-67, CD10, BCL-6 and BCL-2 did not predict prognosis. In conclusion, high Ki-67 expression may predict poor prognosis in PCNSL. The present study was limited by its sample size and short follow-up time. This requires more evidence to further clinical study.