One-Base-Gap Circular Probe-Mediated Dual Amplification for Isothermal Detection of N6-Methyladenosine Modifications.

N6-Methyladenosine (m6A) stands out as the predominant internal modification in mammalian RNA, exerting crucial regulatory functions in the metabolism of mRNA. Currently available methods have been limited by an inability to quantify m6A modification at precise sites. In this work, we screened a Bst 2.0 warm start DNA polymerase with the capability of discriminating m6A from adenosine (A) and developed a robust m6A RNA detection method that enables isothermal and ultrasensitive quantification of m6A RNA at single-base resolution. The detection limit of the assay could reach about 0.02 amol, and the quantitative accuracy of the assay was verified in real cell samples. Furthermore, we applied this assay to single-cell analysis and found that the coefficients of variation of the MALAT1 m6A 2611 site in glioblastoma U251 cells showed over 20% higher than in oligodendrocytes MO3.13 cells. This method provides a highly sensitive analytical tool for site-specific m6A detection and quantification, which is expected to provide a basis for precise disease diagnosis and epigenetic transcriptional regulation.

Depth-assisted calibration on learning-based factorization for a compressive light field display.

Due to the widespread applications of high-dimensional representations in many fields, the three-dimension (3D) display technique is increasingly being used for commercial purpose in a holographic-like and immersive demonstration. However, the visual discomfort and fatigue of 3D head mounts demonstrate the limits of usage in the sphere of marketing. The compressive light field (CLF) display is capable of providing binocular and motion parallaxes by stacking multiple liquid crystal screens without any extra accessories. It leverages optical viewpoint fusion to bring an immersive and visual-pleasing experience for viewers. Unfortunately, its practical application has been limited by processing complexity and reconstruction performance. In this paper, we propose a dual-guided learning-based factorization on polarization-based CLF display with depth-assisted calibration (DAC). This substantially improves the visual performance of factorization in real-time processing. In detail, we first take advantage of a dual-guided network structure under the constraints of reconstructed and viewing images. Additionally, by utilizing the proposed DAC, we distribute each pixel on displayed screens following the real depth. Furthermore, the subjective performance is increased by using a Gauss-distribution-based weighting (GDBW) toward the concentration of the observer's angular position. Experimental results illustrate the improved performance in qualitative and quantitative aspects over other competitive methods. A CLF prototype is assembled to verify the practicality of our factorization.

Combination of RCA and DNAzyme for Dual-Signal Isothermal Amplification of Exosome RNA.

The RNA contained in exosomes plays a crucial role in information transfer between cells in various life activities. The accurate detection of low-abundance exosome RNA (exRNA) is of great significance for cell function studies and the early diagnosis of diseases. However, their intrinsic properties, such as their short length and high sequence homology, represent great challenges for exRNA detection. In this paper, we developed a dual-signal isothermal amplification method based on rolling circle amplification (RCA) coupled with DNAzyme (RCA-DNAzyme). The sensitive detection of low-abundance exRNA, the specific recognition of their targets and the amplification of the detection signal were studied and explored. By designing padlock probes to specifically bind to the target exRNA, while relying on the ligation reaction to enhance recognition, the precise targeting of exosome RNA was realized. The combination of RCA and DNAzyme could achieve a twice-as-large isothermal amplification of the signal compared to RCA alone. This RCA-DNAzyme assay could sensitively detect a target exRNA at a concentration as low as 527 fM and could effectively distinguish the target from other miRNA sequences. In addition, this technology was successfully proven to be effective for the quantitative detection of miR-21 by spike recovery, providing a new research approach for the accurate detection of low-abundance exRNA and the exploration of unknown exRNA functions.

Dual-Functional Nanocluster Probe-Based Single-Cell Analysis of RNA Splice Variants.

Differential expression of RNA splice variants among individual cells accounts for cell heterogeneity of gene expression, which plays a key role in the regulation of the immune system. However, currently available techniques face difficulties in achieving single-cell analysis of RNA splice variants with high base resolution, high spatial resolution and accurate quantification. Herein, we constructed DNA-templated dual-functional nanocluster probes to achieve in situ imaging and accurate quantification of RNA splice variants at the single-cell level. By designing ultrasmall nanocluster labeled probes to directly target the splicing junction sequence of RNA splice variants, the base recognition resolution is significantly improved. Benefit from the controllable fluorescence of nanoclusters, in situ imaging and genotyping of RNA splice variants are achieved. Due to the atom-precise nanocluster, RNA splice variants can be accurately quantified by laser ablation inductively coupled plasma mass spectrometry at the single-cell level. We further applied the probes to explore the function of MyD88 splice variants in mononuclear macrophages under immune activation. This strategy provides a novel single-cell analysis tool for studying the functional diversity of the immune system and splicing-related immune diseases.

One-stop measurement model for fast and accurate tensor display characterization.

Many light field displays are fundamentally different from other displays in that they do not have quantized pixels, quantized angular outputs, or a physical screen position, which can make definitions and characterization problematic. We have determined that it is more appropriate to express the spatial resolution in terms of spatial cutoff frequency rather than a physical distance as in the case of a display with actual quantized pixels. This concept is then extended to also encompass angular resolution. The technique exploits the fact that when spatial resolution of a sinusoidal grating pattern is halved, its contrast ratio is reduced by a known proportion. An improved model, based on an earlier design concept, has been developed. It not only can be used to measure spatial and angular cutoff frequencies, but also can enable comprehensive characterization of the display. This model provides fast, simple measurement with good accuracy. It does not use special equipment or require time-consuming subjective evaluations. Using the model to characterize images in a rapid, accurate manner validates the effectiveness of this technique.

Advancing Imbalanced Domain Adaptation: Cluster-Level Discrepancy Minimization With a Comprehensive Benchmark.

Unsupervised domain adaptation methods have been proposed to tackle the problem of covariate shift by minimizing the distribution discrepancy between the feature embeddings of source domain and target domain. However, the standard evaluation protocols assume that the conditional label distributions of the two domains are invariant, which is usually not consistent with the real-world scenarios such as long-tailed distribution of visual categories. In this article, the imbalanced domain adaptation (IDA) is formulated for a more realistic scenario where both label shift and covariate shift occur between the two domains. Theoretically, when label shift exists, aligning the marginal distributions may result in negative transfer. Therefore, a novel cluster-level discrepancy minimization (CDM) is developed. CDM proposes cross-domain similarity learning to learn tight and discriminative clusters, which are utilized for both feature-level and distribution-level discrepancy minimization, palliating the negative effect of label shift during domain transfer. Theoretical justifications further demonstrate that CDM minimizes the target risk in a progressive manner. To corroborate the effectiveness of CDM, we propose two evaluation protocols according to the real-world situation and benchmark existing domain adaptation approaches. Extensive experiments demonstrate that negative transfer does occur due to label shift, while our approach achieves significant improvement on imbalanced datasets, including Office-31, Image-CLEF, and Office-Home.

Visual genetic typing of glioma using proximity-anchored in situ spectral coding amplification.

Gliomas are histologically and genetically heterogeneous tumors. However, classical histopathological typing often ignores the high heterogeneity of tumors and thus cannot meet the requirements of precise pathological diagnosis. Here, proximity-anchored in situ spectral coding amplification (ProxISCA) is proposed for multiplexed imaging of RNA mutations, enabling visual typing of brain gliomas with different pathological grades at the single-cell and tissue levels. The ligation-based padlock probe can discriminate one-nucleotide variations, and the design of proximity primers enables the anchoring of amplicons on target RNA, thus improving localization accuracy. The DNA module-based spectral coding strategy can dramatically improve the multiplexing capacity for imaging RNA mutations through one-time labelling, with low cost and simple operation. One-target-one-amplicon amplification confers ProxISCA the ability to quantify RNA mutation copy number with single-molecule resolution. Based on this approach, it is found that gliomas with higher malignant grades express more genes with high correlation at the cellular and tissue levels and show greater cellular heterogeneity. ProxISCA provides a tool for glioma research and precise diagnosis, which can reveal the relationship between cellular heterogeneity and glioma occurrence or development and assist in pathological prognosis.

Advancing Image Understanding in Poor Visibility Environments: A Collective Benchmark Study.

Existing enhancement methods are empirically expected to help the high-level end computer vision task: however, that is observed to not always be the case in practice. We focus on object or face detection in poor visibility enhancements caused by bad weathers (haze, rain) and low light conditions. To provide a more thorough examination and fair comparison, we introduce three benchmark sets collected in real-world hazy, rainy, and low-light conditions, respectively, with annotated objects/faces. We launched the UG2+ challenge Track 2 competition in IEEE CVPR 2019, aiming to evoke a comprehensive discussion and exploration about whether and how low-level vision techniques can benefit the high-level automatic visual recognition in various scenarios. To our best knowledge, this is the first and currently largest effort of its kind. Baseline results by cascading existing enhancement and detection models are reported, indicating the highly challenging nature of our new data as well as the large room for further technical innovations. Thanks to a large participation from the research community, we are able to analyze representative team solutions, striving to better identify the strengths and limitations of existing mindsets as well as the future directions.