A crossbar array of magnetoresistive memory devices for in-memory computing.

Implementations of artificial neural networks that borrow analogue techniques could potentially offer low-power alternatives to fully digital approaches1-3. One notable example is in-memory computing based on crossbar arrays of non-volatile memories4-7 that execute, in an analogue manner, multiply-accumulate operations prevalent in artificial neural networks. Various non-volatile memories-including resistive memory8-13, phase-change memory14,15 and flash memory16-19-have been used for such approaches. However, it remains challenging to develop a crossbar array of spin-transfer-torque magnetoresistive random-access memory (MRAM)20-22,  despite the technology's practical advantages such as endurance and large-scale commercialization5. The difficulty stems from the low resistance of MRAM, which would result in large power consumption in a conventional crossbar array that uses current summation for analogue multiply-accumulate operations. Here we report a 64 × 64 crossbar array based on MRAM cells that overcomes the low-resistance issue with an architecture that uses resistance summation for analogue multiply-accumulate operations. The array is integrated with readout electronics in 28-nanometre complementary metal-oxide-semiconductor technology. Using this array, a two-layer perceptron is implemented to classify 10,000 Modified National Institute of Standards and Technology digits with an accuracy of 93.23 per cent (software baseline: 95.24 per cent). In an emulation of a deeper, eight-layer Visual Geometry Group-8 neural network with measured errors, the classification accuracy improves to 98.86 per cent (software baseline: 99.28 per cent). We also use the array to implement a single layer in a ten-layer neural network to realize face detection with an accuracy of 93.4 per cent.

Probiogenomic In-Silico Analysis and Safety Assessment of Lactiplantibacillus plantarum DJF10 Strain Isolated from Korean Raw Milk.

The whole genome sequence of Lactiplantibacillus plantarum DJF10, isolated from Korean raw milk, is reported, along with its genomic analysis of probiotics and safety features. The genome consists of 29 contigs with a total length of 3,385,113 bp and a GC content of 44.3%. The average nucleotide identity and whole genome phylogenetic analysis showed the strain belongs to Lactiplantibacillus plantarum with 99% identity. Genome annotation using Prokka predicted a total of 3235 genes, including 3168 protein-coding sequences (CDS), 59 tRNAs, 7 rRNAs and 1 tmRNA. The functional annotation results by EggNOG and KEGG showed a high number of genes associated with genetic information and processing, transport and metabolism, suggesting the strain's ability to adapt to several environments. Various genes conferring probiotic characteristics, including genes related to stress adaptation to the gastrointestinal tract, biosynthesis of vitamins, cell adhesion and production of bacteriocins, were identified. The CAZyme analysis detected 98 genes distributed under five CAZymes classes. In addition, several genes encoding carbohydrate transport and metabolism were identified. The genome also revealed the presence of insertion sequences, genomic islands, phage regions, CRISPR-cas regions, and the absence of virulence and toxin genes. However, the presence of hemolysin and antibiotic-resistance-related genes detected in the KEGG search needs further experimental validation to confirm the safety of the strain. The presence of two bacteriocin clusters, sactipeptide and plantaricin J, as detected by the BAGEL 4 webserver, confer the higher antimicrobial potential of DJF10. Altogether, the analyses in this study performed highlight this strain's functional characteristics. However, further in vitro and in vivo studies are required on the safety assurance and potential application of L. plantarum DJF10 as a probiotic agent.

Daily Supplementation with Bifidobacterium longum KACC91563 Alleviates Allergic Contact Dermatitis in an Animal Model.

Allergic contact dermatitis (ACD) is the most common chronic inflammatory skin disease (or immune-mediated disease), causing disruption to our psychological condition and life quality. In this study, the therapeutic properties of probiotic Bifidobacterium longum (B. longum) was investigated by using an ACD-induced animal model. For ACD induction, BALB/c mice ear and dorsal skin were sensitized with 240 µL of 1% (w/v) 2,4-dinitrochlorobenzene (DNCB) twice (3-day intervals). After a week of the first induction, the mice were re-sensitized by painting on their dorsal skin and ear with 0.4% (w/v) DNCB for a further three times (once per week). Before the ACD induction of 2 weeks and throughout the trial period, the BALB/c mice were supplemented daily with 1 mL of 1.0 × 109 CFU or 5.0 × 109 CFU B. longum using an intragastric gavage method. The ACD-induced mice without B. longum supplementation were used as a control. Results show that B. longum supplementation significantly alleviated ACD symptoms (e.g., ear swelling, epidermal damage) and immune response (e.g., reduced immune cell recruitment, serum IgE level, and cytokine production). The therapeutic efficiency of B. longum increased as the supplementation dose increased. Thus, daily supplementation with 5.0 × 109 CFU probiotic B. longum could be an effective method for the prevention and treatment of ACD.

Characterization of a frozen shoulder model using immobilization in rats.

BACKGROUND: The objective of this study was to investigate serial changes for histology of joint capsule and range of motion of the glenohumeral joint after immobilization in rats. We hypothesized that a rat shoulder contracture model using immobilization would be capable of producing effects on the glenohumeral joint similar to those seen in patients with frozen shoulder. METHODS: Sixty-four Sprague-Dawley rats were randomly divided into one control group (n = 8) and seven immobilization groups (n = 8 per group) that were immobilized with molding plaster for 3 days, or for 1, 2, 3, 4, 5, or 6 weeks. At each time point, eight rats were euthanized for histologic evaluation of the axillary recess and for measurement of the abduction angle. RESULTS: Infiltration of inflammatory cells was found in the synovial tissue until 2 weeks after immobilization. However, inflammatory cells were diminished and fibrosis was dominantly observed in the synovium and subsynovial tissue 3 weeks after immobilization. From 1 week after immobilization, the abduction angle of all immobilization groups at each time point was significantly lower than that of the control group. CONCLUSIONS: Our study demonstrated that a rat frozen shoulder model using immobilization generates the pathophysiologic process of inflammation leading to fibrosis on the glenohumeral joint similar to that seen in patients with frozen shoulder. This model was attained within 3 weeks after immobilization. It may serve as a useful tool to investigate pathogenesis at the molecular level and identify potential target genes that are involved in the development of frozen shoulder.

Improving fatty acid availability for bio-hydrocarbon production in Escherichia coli by metabolic engineering.

Previous studies have demonstrated the feasibility of producing fatty-acid-derived hydrocarbons in Escherichia coli. However, product titers and yields remain low. In this work, we demonstrate new methods for improving fatty acid production by modifying central carbon metabolism and storing fatty acids in triacylglycerol. Based on suggestions from a computational model, we deleted seven genes involved in aerobic respiration, mixed-acid fermentation, and glyoxylate bypass (in the order of cyoA, nuoA, ndh, adhE, dld, pta, and iclR) to modify the central carbon metabolic/regulatory networks. These gene deletions led to increased total fatty acids, which were the highest in the mutants containing five or six gene knockouts. Additionally, when two key enzymes in the fatty acid biosynthesis pathway were over-expressed, we observed further increase in strain △cyoA△adhE△nuoA△ndh△pta△dld, leading to 202 mg/g dry cell weight of total fatty acids, ~250% of that in the wild-type strain. Meanwhile, we successfully introduced a triacylglycerol biosynthesis pathway into E. coli through heterologous expression of wax ester synthase/acyl-coenzyme:diacylglycerol acyltransferase (WS/DGAT) enzymes. The added pathway improved both the amount and fuel quality of the fatty acids. These new metabolic engineering strategies are providing promising directions for future investigation.