Search

In this work, a novel synaptic array architecture enabling forward propagation (FP) and backward propagation (BP) in the NAND flash memory is proposed for the first time for on-chip learning. In the proposed synaptic architecture, positive synaptic weight and negative synaptic weight are separated in different arrays to enable weights to be transposed correctly. In addition, source-lines (SLs) are separated, which is different from the conventional NAND flash memory, to enable both the FP and BP in the NAND flash memory. By applying input and error input to bitlines (BLs) and string-select lines (SSLs) in NAND cell array, respectively, accurate vector–matrix multiplication is successfully carried out in both FP and BP eliminating the effect of pass cells. At a read voltage of 2 V, the inference accuracy of 95.58% which is comparable to that of 95.81% obtained with perfect linear device is achieved. The proposed on-chip learning system is much more robust to weight variation compared to the off-chip learning system. Finally, superiority of the proposed on-chip learning architecture is verified by circuit simulation of a neural network.

A novel design method using unsigned input is proposed for a high-density and parallel synaptic string architecture capable of bit-wise operation and bit-counting utilizing NAND flash memory. Though the NAND flash memory has a cell string structure, unsigned binary input enables analogue and parallel bit-counting in NAND flash memory, while achieving high accuracy comparable to that of signed input. Adopting unsigned input allows using current sense amplifier as neuron circuit, which reduces burden of peripheral circuits. In addition, the operation scheme for convolution layers is proposed for parallel convolution operation utilizing NAND flash memory. We show that sufficiently low device variation of 7.2 % obtained by applying 1 erase or program pulse achieves accuracy of 98.12 % and 87.12 % for MNIST and CIFAR 10 patterns, respectively.

To effectively control the spread of new infectious diseases, there is a need for highly sensitive diagnostic methods to detect viral nucleic acids rapidly. This study outlines a universal and simple detection strategy that uses magnetic nanoparticles (MNPs) and a novel MagR-MazE fusion protein for molecular diagnostics to facilitate sensitive detection. This study has engineered a novel MNP conjugate that can be generated easily, without using many chemical reagents. The technique is a nucleic acid detection method, using MagR-MazE fusion protein-conjugated MNPs, where the results can be visualized with the naked eye, regardless of the oligonucleotide sequences of the target in the lateral flow assay. This method could sensitively detect polymerase chain reaction (PCR) products of 16S ribosomal RNA (rRNA) and the 2019-nCoV-N-positive control gene in 5 min. It shows a low limit of detection (LoD) of 0.013 ng/μL for dsDNA. It is simpler and more rapid, sensitive, and versatile than other techniques, making it suitable for point-of-care testing. The proposed detection system and MNP conjugation strategy using a fusion protein can be widely applied to various fields requiring rapid on-site diagnosis.

Most prevalent infectious diseases worldwide are caused by mediators such as insects and characterized by high mortality and morbidity, thereby creating a global public health concern. Therefore, a sensitive, selective detection platform for diagnosing diseases in the early stages of infection is needed to prevent disease spread and to protect public health. Here, we developed novel DNA aptamers specific to the nucleocapsid protein (NP) of the severe fever with thrombocytopenia syndrome (SFTS) virus and synthesized ssDNA-binding protein-conjugated liposomes encapsulated with horseradish peroxidase (HRP) for application in a simple and universal platform. This platform achieved highly sensitive detection of the NP by measuring the colorimetric signal following lysis of the HRP encapsulated liposomes, mediated by a mixture of 3,3',5,5'-tetramethylbenzidine and H2O2 solution. The limit of detection was 0.009 ng·mL-1, and NP was successfully detected in diluted human serum with a high recovery rate. Moreover, this method was specific and did not exhibit cross-reactivity among NPs of other virus types. These results demonstrated the efficacy of the proposed method as a highly sensitive, specific, and universal diagnostic tool for potential application in monitoring of the early stages of infectious diseases.

To facilitate the exact diagnosis of multiple influenza virus types that are reported each year, the quantitative, sandwich-format enzyme-linked immunosorbent assay (ELISA) is widely used. In this method, enzyme-linked secondary antibodies or aptamers serve as detection reagents for signal amplification or colorimetric analysis. Aptamers have similar target affinities and specificities to antibodies, and they are relatively easy to label and can be applied to a variety of methods. For this reason, many quantitative analysis methods have recently been proposed that have a heterogeneous sandwich format and use an antibody with high specificity for a target and an aptamer that can be easily labeled for signal amplification. Aptamers obtained through the conventional systematic evolution of ligands by the exponential enrichment (SELEX) method have a limitation in the heterogeneous sandwich format assay because they often bind the antibody that acts as the capturing agent. In this paper, we introduce an effective method for selecting aptamers that increases the signal-to-noise ratio in a heterogenous sandwich-type immunosensor and confirms the efficiency of the selected aptamer candidates in the colorimetric assay. Using the proposed method, four aptamer candidates with Kd values ranging from 77.6 to 125.7 nM were obtained. Of them, INFA-apt4 showed the strongest binding affinity, with quantitative detection of nucleoprotein ranging from 1 ng mL−1 to 300 ng mL−1 and a minimum detection limit of 1.02 ng mL−1, which was 10-fold more sensitive than the values offered by conventional antibody-based methods.

The detection of trace protein biomarkers is essential in the diagnostic field. Protein detection systems ranging from widely used enzyme-linked immunosorbent assays to simple, inexpensive approaches, such as lateral flow immunoassays, play critical roles in medical and drug research. Despite continuous progress, current systems are insufficient for the diagnosis of diseases that require high sensitivity. In this study, we developed a heterogeneous sandwich-type sensing platform based on recombinase polymerase amplification using DNA aptamers specific to the target biomarker. Only the DNA bound to the target in the form of a heterogeneous sandwich was selectively amplified, and the fluorescence signal of an intercalating dye added before the amplification reaction was detected, thereby enabling high specificity and sensitivity. We applied this method for the detection of protein biomarkers for various infectious diseases including severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and observed attomolar-level detection of biomarkers and low cross-reactivity between different viruses. We also confirmed detection efficiency of the proposed method using clinical samples. These results demonstrate that the proposed sensing platform can be used to diagnose various diseases requiring high sensitivity, specificity, and accuracy.

A simple, universal, and sensitive colorimetric biosensor for detecting of various biomarkers was devised using a target-specific DNA aptamer, as the recognition element, and engineered with streptavidin-fusion replication protein A 70 kDa (RPA70A) linked to biotin-horseradish peroxidase, as the colorimetric element. To improve sensitivity and stability compared to other colorimetric sensing platforms, we developed a novel detection strategy by integrating a newly selected heterogeneous sandwich DNA aptamer and protein engineering in this study. The proposed method is based on a change in color from colorless to blue due to the interaction of the aptamer with RPA70A in the presence of the target; this color change could be observed by the naked eye or measured with a UV–vis spectrometer. We confirmed its high sensitivity and specificity for two model targets using their aptamers under optimal experimental conditions. In addition, the feasibility of the assay was investigated in clinical samples containing NPs of influenza A or B virus. These results suggest that our detection system developed herein can be universally applied to the diagnosis of various diseases owing to its stability, sensitivity, and specificity.

Werner syndrome protein (WRN) and Fanconi anemia group J protein (FANCJ) are human DNA helicases that contribute to genome maintenance. They interact with replication protein A (RPA), and these interactions dramatically enhance the unwinding activities of both helicases. Even though the interplay between these helicases and RPA is particularly important in the chemoresistance pathway of cancer cells, the precise binding regions, interfaces, and properties have not yet been characterized. Here we present systematic NMR analyses and fluorescence polarization anisotropy assays of both helicase-RPA interactions for defining core binding regions and binding affinities. Our results showed that two acidic repeats of human WRN bind to RPA70N and RPA70A. For FANCJ, the acidic-rich sequence in the C-terminal domain is the binding region for RPA70N. Our results suggest that each helicase interaction has unique features, although they both fit an acidic peptide into a basic cleft for RPA binding. Our findings shed light on the protein interactions involved in overcoming the DNA-damaging agents employed in the treatment of cancer and thus potentially provide insight into enhancing the efficacy of cancer therapy.

Paper-based lateral flow immunoassays (LFIAs) using conventional sandwich-type immunoassays are one of the most commonly used point-of-care (PoC) tests. However, the application of gold nanoparticles (AuNPs) in LFIAs does not meet sensitivity requirements for the detection of infectious diseases or biomarkers present at low concentrations in body fluids because of the limited number of AuNPs that can bind to the target. To overcome this problem, we first developed a single-stranded DNA binding protein (RPA70A, DNA binding domain A of human Replication Protein A 70 kDa) conjugated to AuNPs for a sandwich assay using a capture antibody immobilized in the LFIA and an aptamer as a detection probe, thus, enabling signal intensity enhancement by attaching several AuNPs per aptamer. We applied this method to detect the influenza nucleoprotein (NP) and cardiac troponin I (cTnI). We visually detected spiked targets at a low femtomolar range, with limits of detection for NP in human nasal fluid and for cTnI in serum of 0.26 and 0.23 pg·mL-1, respectively. This technique showed significantly higher sensitivity than conventional methods that are widely used in LFIAs involving antibody-conjugated AuNPs. These results suggest that the proposed method can be universally applied to the detection of substances requiring high sensitivity and can be used in the field of PoC testing for early disease diagnosis.

A simple and rapid As3+ detection method using 3-nitro-L-tyrosine (N-Tyr) is reported. We discovered the specific property of N-Tyr, which specifically chelates As3+. The reaction between As3+ and N-Tyr induces a prompt color change to vivid yellow, concomitantly increasing the absorbance at 430 nm. The selectivity for As3+ is confirmed by competitive binding experiments with various metal ions (Hg2+, Pb2+, Cd2+, Cr3+, Mg2+, Ni2+, Cu2+, Fe2+, Ca2+, Zn2+, and Mn2+). Also, the N-Tyr binding site, binding affinity, and As3+/N-Tyr reaction stoichiometry are investigated. The specific reaction is utilized to design a sensor that enables the quantitative detection of As3+ in the 0.1&ndash, 100 &mu, M range with good linearity (R2 = 0.995). Furthermore, the method&rsquo, s applicability for the analysis of real samples, e.g., tap and river water, is successfully confirmed, with good recoveries (94.32&ndash, 109.15%) using As3+-spiked real water samples. We believe that our discovering and its application for As3+ analysis can be effectively utilized in environmental analyses such as those conducted in water management facilities, with simplicity, rapidity, and ease.