Deep learning-based activity recognition and fine motor identification using 2D skeletons of cynomolgus monkeys.

Video-based action recognition is becoming a vital tool in clinical research and neuroscientific study for disorder detection and prediction. However, action recognition currently used in non-human primate (NHP) research relies heavily on intense manual labor and lacks standardized assessment. In this work, we established two standard benchmark datasets of NHPs in the laboratory: MonkeyinLab (MiL), which includes 13 categories of actions and postures, and MiL2D, which includes sequences of two-dimensional (2D) skeleton features. Furthermore, based on recent methodological advances in deep learning and skeleton visualization, we introduced the MonkeyMonitorKit (MonKit) toolbox for automatic action recognition, posture estimation, and identification of fine motor activity in monkeys. Using the datasets and MonKit, we evaluated the daily behaviors of wild-type cynomolgus monkeys within their home cages and experimental environments and compared these observations with the behaviors exhibited by cynomolgus monkeys possessing mutations in the MECP2 gene as a disease model of Rett syndrome (RTT). MonKit was used to assess motor function, stereotyped behaviors, and depressive phenotypes, with the outcomes compared with human manual detection. MonKit established consistent criteria for identifying behavior in NHPs with high accuracy and efficiency, thus providing a novel and comprehensive tool for assessing phenotypic behavior in monkeys.

Abundance of Transgene Transcript Variants Associated with Somatically Active Transgenic Helitrons from Multiple T-DNA Integration Sites in Maize.

Helitrons, a novel type of mysterious DNA transposons discovered computationally prior to bench work confirmation, are components ubiquitous in most sequenced genomes of various eukaryotes, including plants, animals, and fungi. There is a paucity of empirical evidence to elucidate the mechanism of Helitrons transposition in plants. Here, by constructing several artificial defective Helitron (dHel) reporter systems, we aim to identify the autonomous Helitrons (aHel) in maize genetically and to demonstrate the transposition and repair mechanisms of Helitrons upon the dHel-GFP excision in maize. When crossing with various inbred lines, several transgenic lines produced progeny of segregated, purple-blotched kernels, resulting from a leaky expression of the C1 gene driven by the dHel-interrupted promoter. Transcription analysis indicated that the insertion of different dHels into the C1 promoter or exon would lead to multiple distinct mRNA transcripts corresponding to transgenes in the host genome. Simple excision products and circular intermediates of dHel-GFP transposition have been detected from the leaf tissue of the seedlings in F1 hybrids of transgenic lines with corresponding c1 tester, although they failed to be detected in all primary transgenic lines. These results revealed the transposition and repair mechanism of Helitrons in maize. It is strongly suggested that this reporter system can detect the genetic activity of autonomic Helitron at the molecular level. Sequence features of dHel itself, together with the flanking regions, impact the excision activity of dHel and the regulation of the dHel on the transcription level of the host gene.

Cloning of Maize TED Transposon into Escherichia coli Reveals the Polychromatic Sequence Landscape of Refractorily Propagated Plasmids.

MuDR, the founder member of the Mutator superfamily and its MURA transcripts, has been identified as toxic sequences to Escherichia coli (E. coli), which heavily hindered the elucidation of the biochemical features of MURA transposase and confined the broader application of the Mutator system in other organisms. To harness less constrained systems as alternatives, we attempted to clone TED and Jittery, two recently isolated autonomous Mutator-like elements (MULEs) from maize, respectively. Their full-length transcripts and genomic copies are successfully cloned when the incubation time for bacteria to recover from heat shock is extended appropriately prior to plating. However, during their proliferation in E. coli, TED transformed plasmids are unstable, as evidenced by derivatives from which frameshift, deletion mutations, or IS transposon insertions are readily detected. Our results suggest that neither leaky expression of the transposase nor the presence of terminal inverse repeats (TIRs) are responsible for the cloning barriers, which were once ascribed to the presence of the Shine-Dalgarno-like sequence. Instead, the internal sequence of TED (from 1250 to 2845 bp), especially the exons in this region, was the most likely causer. The findings provide novel insights into the property and function of the Mutator superfamily and shed light on the dissection of toxic effects on cloning from MULEs.

A Heterogeneous RISC-V Processor for Efficient DNN Application in Smart Sensing System.

Extracting features from sensing data on edge devices is a challenging application for which deep neural networks (DNN) have shown promising results. Unfortunately, the general micro-controller-class processors which are widely used in sensing system fail to achieve real-time inference. Accelerating the compute-intensive DNN inference is, therefore, of utmost importance. As the physical limitation of sensing devices, the design of processor needs to meet the balanced performance metrics, including low power consumption, low latency, and flexible configuration. In this paper, we proposed a lightweight pipeline integrated deep learning architecture, which is compatible with open-source RISC-V instructions. The dataflow of DNN is organized by the very long instruction word (VLIW) pipeline. It combines with the proposed special intelligent enhanced instructions and the single instruction multiple data (SIMD) parallel processing unit. Experimental results show that total power consumption is about 411 mw and the power efficiency is about 320.7 GOPS/W.

Efficient Resource-Aware Convolutional Neural Architecture Search for Edge Computing with Pareto-Bayesian Optimization.

With the development of deep learning technologies and edge computing, the combination of them can make artificial intelligence ubiquitous. Due to the constrained computation resources of the edge device, the research in the field of on-device deep learning not only focuses on the model accuracy but also on the model efficiency, for example, inference latency. There are many attempts to optimize the existing deep learning models for the purpose of deploying them on the edge devices that meet specific application requirements while maintaining high accuracy. Such work not only requires professional knowledge but also needs a lot of experiments, which limits the customization of neural networks for varied devices and application scenarios. In order to reduce the human intervention in designing and optimizing the neural network structure, multi-objective neural architecture search methods that can automatically search for neural networks featured with high accuracy and can satisfy certain hardware performance requirements are proposed. However, the current methods commonly set accuracy and inference latency as the performance indicator during the search process, and sample numerous network structures to obtain the required neural network. Lacking regulation to the search direction with the search objectives will generate a large number of useless networks during the search process, which influences the search efficiency to a great extent. Therefore, in this paper, an efficient resource-aware search method is proposed. Firstly, the network inference consumption profiling model for any specific device is established, and it can help us directly obtain the resource consumption of each operation in the network structure and the inference latency of the entire sampled network. Next, on the basis of the Bayesian search, a resource-aware Pareto Bayesian search is proposed. Accuracy and inference latency are set as the constraints to regulate the search direction. With a clearer search direction, the overall search efficiency will be improved. Furthermore, cell-based structure and lightweight operation are applied to optimize the search space for further enhancing the search efficiency. The experimental results demonstrate that with our method, the inference latency of the searched network structure reduced 94.71% without scarifying the accuracy. At the same time, the search efficiency increased by 18.18%.

Intranasal immunization with recombinant chlamydial protease-like activity factor attenuates atherosclerotic pathology following Chlamydia pneumoniae infection in mice.

We have shown previously that intranasal vaccination with recombinant chlamydial protease-like activity factor (rCPAF: antigen) and interleukin-12 (IL-12) as an adjuvant induces robust protection against pathological consequences of female genital tract infection with Chlamydia muridarum, a closely related species and a rodent model for the human pathogen Chlamydia trachomatis. Another related species Chlamydia pneumoniae, a human respiratory pathogen, has been associated with exacerbation of atherosclerotic pathology. CPAF is highly conserved among Chlamydia spp. leading us to hypothesize that immunization with rCPAF with IL-12 will protect against high-fat diet (HFD) and C. pneumoniae-induced acceleration of atherosclerosis. rCPAF ± IL-12 immunization induced robust splenic antigen (Ag)-specific IFN-γ and TNF-α production and significantly elevated serum total anti-CPAF Ab, IgG2c, and IgG1 antibody levels compared to mock or IL-12 alone groups. The addition of IL-12 to rCPAF significantly elevated splenic Ag-specific IFN-γ production and IgG2c/IgG1 anti-CPAF antibody ratio. Following intranasal C. pneumoniae challenge and HFD feeding, rCPAF ± IL-12-immunized mice displayed significantly enhanced splenic IFN-γ, not TNF-α, response on days 6 and 9 after challenge, and significantly reduced lung chlamydial burden on day 9 post-challenge compared to mock- or IL-12-immunized mice. Importantly, rCPAF ± IL-12-immunized mice displayed significantly reduced atherosclerotic pathology in the aortas after C. pneumoniae challenge. Serum cholesterol levels were comparable between the groups suggesting that the observed differences in pathology were due to protective immunity against the infection. Together, these results confirm and extend our previous observations that CPAF is a promising candidate antigen for a multisubunit vaccine regimen to protect against Chlamydia-induced pathologies, including atherosclerosis.

Evaluation of LL-37 antimicrobial peptide derivatives alone and in combination with vancomycin against S. aureus.

Treatment of Staphylococcus aureus infections continues to be a challenge due to antimicrobial resistance. Endogenous antimicrobial peptides may offer a new option for treating S. aureus infections but several factors limit their clinical utility. Herein, we studied the activity of the antimicrobial peptide LL-37 and two truncated derivatives, LL-13 and LL-17 alone and in combination with vancomycin against a range of drug-resistant S. aureus strains including methicillin resistant S. aureus (MRSA) and vancomycin resistant S. aureus (VRSA) strains in vitro. When used with vancomycin, LL-13 and LL-17 displayed synergy against VRSA and showed the ability to restore sensitivity to vancomycin after pretreatment. In addition, LL-13 and LL-17 showed a strong ability to inhibit S. aureus biofilm production. LL-37 derivatives may be useful in treating infections that are resistant to vancomycin or in scenarios where biofilm formation is a concern.

RNA-guided Cas9 as an in vivo desired-target mutator in maize.

The RNA-guided Cas9 system is a versatile tool for genome editing. Here, we established a RNA-guided endonuclease (RGEN) system as an in vivo desired-target mutator (DTM) in maize to reduce the linkage drag during breeding procedure, using the LIGULELESS1 (LG1) locus as a proof-of-concept. Our system showed 51.5%-91.2% mutation frequency in T0 transgenic plants. We then crossed the T1 plants stably expressing DTM with six diverse recipient maize lines and found that 11.79%-28.71% of the plants tested were mutants induced by the DTM effect. Analysis of successive F2 plants indicated that the mutations induced by the DTM effect were largely heritable. Moreover, DTM-generated hybrids had significantly smaller leaf angles that were reduced more than 50% when compared with that of the wild type. Planting experiments showed that DTM-generated maize plants can be grown with significantly higher density and hence greater yield potential. Our work demonstrate that stably expressed RGEN could be implemented as an in vivoDTM to rapidly generate and spread desired mutations in maize through hybridization and subsequent backcrossing, and hence bypassing the linkage drag effect in convention introgression methodology. This proof-of-concept experiment can be a potentially much more efficient breeding strategy in crops employing the RNA-guided Cas9 genome editing.