Search

The model weights of SpliceBERT ("models.tar.gz") and data ("data.tar.gz") for running the analysis in manuscript "Self-supervised learning on millions of pre-mRNA sequences from 72 vertebrates improves sequence-based RNA splicing prediction". See https://github.com/biomed-AI/SpliceBERT for the codes.

In this study, anatase TiO2-supported cerium, manganese, and ruthenium mixed oxides (CeOx–MnOx–RuOx/TiO2; CMRT catalysts) were synthesized at different calcination temperatures via conventional impregnation methods and used for selective catalytic reduction (SCR) of NOx with NH3. The effect of calcination temperature on the structure, redox properties, activation performance, surface-acidity properties, and catalytic properties of the CMRT catalysts was investigated. The results show that the CMRT catalyst calcined at 350 °C exhibits the most efficient low-temperature (200 °C, which leads to a decrease in the N2 selectivity of the CMRT catalysts and further results in an increase in the production of NO and N2O byproducts. X-ray photoelectron spectroscopy and in situ diffuse reflectance infrared Fourier transform spectroscopy show that the CMRT catalyst calcined at 350 °C contains more Ce4+, Mn4+, Ru4+, and lattice oxygen, which greatly improve the catalyst’s ability to activate NO that promotes the NH3-SCR reaction. The Run+ sites of the CMRT catalyst calcined at 250 °C are the competitive adsorption sites of NO and NH3 molecules, while those of the CMRT catalyst calcined at 350 and 450 °C are active sites. Both the Langmuir–Hinshelwood (L–H) mechanism and the Eley–Rideal (E–R) mechanism occur on the surface of the CMRT catalyst at the low reaction temperature (100 °C).

This paper reported synergistic bimetallic AuCu/TiO2 catalysts for selective oxidation of 1,2-propanediol to lactic acid under very mild conditions (T 92%). Surface characterization reveals that while AuCu forms alloy structure with larger particle size based on TEM images, strong interaction between Au and Cu species is critical for performance enhancement. According to experimental studies on the influence of 1,2-propanediol and NaOH concentration on reaction rates, it is highly possible that 1,2-propanediol is reacted following single-site Langmuir–Hinshelwood mechanism while NaOH acts as a promoter and may block surface sites under relatively higher concentration.

Funder: National Natural Science Foundation of China (National Science Foundation of China); doi: https://doi.org/10.13039/501100001809, Funder: RCUK | Engineering and Physical Sciences Research Council (EPSRC); doi: https://doi.org/10.13039/501100000266, For a light-emitting diode (LED) to generate light, the minimum voltage required is widely considered to be the emitter's bandgap divided by the elementary charge. Here we show for many classes of LEDs, including those based on perovskite, organic, quantum-dot and III-V semiconductors, light emission can be observed at record-low voltages of 36-60% of their bandgaps, exhibiting a large apparent energy gain of 0.6-1.4 eV per photon. For 17 types of LEDs with different modes of charge injection and recombination (dark saturation currents of ~10-39-10-15 mA cm-2), their emission intensity-voltage curves under low voltages show similar behaviours. These observations and their consistency with the diode simulations suggest the ultralow-voltage electroluminescence arises from a universal origin-the radiative recombination of non-thermal-equilibrium band-edge carriers whose populations are determined by the Fermi-Dirac function perturbed by a small external bias. These results indicate the potential of low-voltage LEDs for communications, computational and energy applications.

Visual saliency plays an important role in many applications, such as image/video retargeting, automatic photo composition, vision-based navigation, etc. Visual saliency can guide these applications to only focus on the important regions, thus reduce the complexity of scene analysis. However, current saliency detection methods generate saliency maps with low resolution or quality, which may not satisfy the requirements of some applications. Moreover, compared with a large amount of research efforts on static images, the saliency models for videos are less well established. In this thesis, we study and propose several models to detect salient objects or regions in images and videos. To address the low resolution problem of saliency maps, we improve the current clustering framework by introducing a two-level clustering strategy based on the complexity of images. We first use the adaptive mean shift algorithm to extract superpixels from the input image, then employ the Gaussian Mixture Model (GMM) to group superpixels based on their appearance similarity. The saliency value is finally calculated for each cluster using compactness metric together with modified PageRank propagation. With the superpixel representation and saliency refinement, this region-based method represents the input image in a perceptually meaningful way and highlights salient regions with full resolution and well-defined boundary. The application of our saliency maps in object recognition shows the potential of the proposed method. For video saliency detection, motivated by the psychological findings that human visual system is extremely sensitive to isolated abrupt stimulus and relative movement, we formulate the saliency detection problem as an unified feature reconstruction problem. For temporal saliency, we use patches in neighboring frames to sparsely reconstruct the target patch in the current frame. We measure the temporal saliency of a patch based on its abruptness, which is estimated by the reconstruction error as well as regularizer, and its motion contrast calculated as the difference of reconstruction coefficients. For spatial saliency, we use the surrounding patches in the same frame to sparsely reconstruct the center patch. The reconstruction error and regularizer are used to measure the local center-surround contrast for spatial saliency detection. The excellent performance of our feature reconstruction in both image and video evaluations justifies the plausibility of feature reconstruction as an explanation for visual saliency. The sparse and low-rank representation demonstrates great potential in subspace learning. For different camera motion, we develop different video saliency detection models based on this powerful technique. With respect to moderate camera motion, we jointly estimate the salient foreground motion and the camera motion via robust alignment with sparse and low-rank decomposition. Consecutive frames are transformed and aligned, and then decomposed to a low-rank matrix representing the background and a sparse matrix indicating the objects with salient motion. We also incorporate useful spatial information including global rarity, local center-surround contrast and location priority, into our model to comprehensively detect spatiotemporal saliency. With regards to large camera motion, our alignment-based model may fail to detect moving objects. Thus we propose to use trajectory representation in the sparse and low-rank decomposition for videos with large camera motion. Under the assumption of orthographic projection, the trajectories from background lie in a subspace spanned by three basis trajectories, i.e. the rank of the background matrix is 3. We estimate the compact background model based on this rank constraint. Furthermore, to enforce the spatial connectivity and motion coherency constraint, a Markov Random Field (MRF) is built for foreground estimation. This model is evaluated on a set of challenging sequences and shows superior performance compared to several state-of-the-art methods. Doctor of Philosophy (SCE)