GraphPro: An interpretable graph neural network-based model for identifying promoters in multiple species.

Promoters are DNA sequences that bind with RNA polymerase to initiate transcription, regulating this process through interactions with transcription factors. Accurate identification of promoters is crucial for understanding gene expression regulation mechanisms and developing therapeutic approaches for various diseases. However, experimental techniques for promoter identification are often expensive, time-consuming, and inefficient, necessitating the development of accurate and efficient computational models for this task. Enhancing the model's ability to recognize promoters across multiple species and improving its interpretability pose significant challenges. In this study, we introduce a novel interpretable model based on graph neural networks, named GraphPro, for multi-species promoter identification. Initially, we encode the sequences using k-tuple nucleotide frequency pattern, dinucleotide physicochemical properties, and dna2vec. Subsequently, we construct two feature extraction modules based on convolutional neural networks and graph neural networks. These modules aim to extract specific motifs from the promoters, learn their dependencies, and capture the underlying structural features of the promoters, providing a more comprehensive representation. Finally, a fully connected neural network predicts whether the input sequence is a promoter. We conducted extensive experiments on promoter datasets from eight species, including Human, Mouse, and Escherichia coli. The experimental results show that the average Sn, Sp, Acc and MCC values of GraphPro are 0.9123, 0.9482, 0.8840 and 0.7984, respectively. Compared with previous promoter identification methods, GraphPro not only achieves better recognition accuracy on multiple species, but also outperforms all previous methods in cross-species prediction ability. Furthermore, by visualizing GraphPro's decision process and analyzing the sequences matching the transcription factor binding motifs captured by the model, we validate its significant advantages in biological interpretability. The source code for GraphPro is available at https://github.com/liuliwei1980/GraphPro.

MMD-DTA: A multi-modal deep learning framework for drug-target binding affinity and binding region prediction.

The prediction of drug-target affinity (DTA) plays a crucial role in drug development and the identification of potential drug targets. In recent years, computer-assisted DTA prediction has emerged as a significant approach in this field. In this study, we propose a multi-modal deep learning framework called MMD-DTA for predicting drug-target binding affinity and binding regions. The model can predict DTA while simultaneously learning the binding regions of drug-target interactions through unsupervised learning. To achieve this, MMD-DTA first uses graph neural networks and target structural feature extraction network to extract multi-modal information from the sequences and structures of drugs and targets. It then utilizes the feature interaction and fusion modules to generate interaction descriptors for predicting DTA and interaction strength for binding region prediction. Our experimental results demonstrate that MMD-DTA outperforms existing models based on key evaluation metrics. Furthermore, external validation results indicate that MMD-DTA enhances the generalization capability of the model by integrating sequence and structural information of drugs and targets. The model trained on the benchmark dataset can effectively generalize to independent virtual screening tasks. The visualization of drug-target binding region prediction showcases the interpretability of MMD-DTA, providing valuable insights into the functional regions of drug molecules that interact with proteins.

SSCRB: Predicting circRNA-RBP Interaction Sites Using a Sequence and Structural Feature-Based Attention Model.

The prediction of interaction sites between circular RNA (circRNA) and RNA binding proteins (RBPs) is crucial for regulating diseases and discovering new treatment approaches. Computational models have been widely used to predict circRNA-RBP interaction sites due to the availability of genome-wide circRNA binding event data. However, efficiently obtaining multi-scale circRNA features to improve prediction accuracy remains a challenging problem. In this study, we propose SSCRB, a lightweight model for predicting circRNA-RBP interaction sites. Our model extracts both sequence and structural features of circRNA and incorporates multi-scale features through the attention mechanism. Furthermore, we develop an ensemble model by combining multiple submodels to enhance predictive performance and generalizability. We evaluate SSCRB on 37 circRNA datasets and compare it with other state-of-the-art methods. The average AUC of SSCRB is 97.66%, demonstrating its efficiency and robustness. SSCRB outperforms other methods in terms of prediction accuracy while requiring significantly fewer computational resources.

Quantitative analysis of drug tablet aging by fast hyper-spectral stimulated Raman scattering microscopy.

Pharmaceutical development of solid-state formulations requires testing active pharmaceutical ingredients (API) and excipients for uniformity and stability. Solid-state properties such as component distribution and grain size are crucial factors that influence the dissolution profile, which greatly affect drug efficacy and toxicity, and can only be analyzed spatially by chemical imaging (CI) techniques. Current CI techniques such as near infrared microscopy and confocal Raman spectroscopy are capable of high chemical and spatial resolution but cannot achieve the measurement speeds necessary for integration into the pharmaceutical production and quality assurance processes. To fill this gap, we demonstrate fast chemical imaging by epi-detected sparse spectral sampling stimulated Raman scattering to quantify API and excipient degradation and distribution.

A Biological Feature and Heterogeneous Network Representation Learning-Based Framework for Drug-Target Interaction Prediction.

The prediction of drug-target interaction (DTI) is crucial to drug discovery. Although the interactions between the drug and target can be accurately verified by traditional biochemical experiments, the determination of DTI through biochemical experiments is a time-consuming, laborious, and expensive process. Therefore, we propose a learning-based framework named BG-DTI for drug-target interaction prediction. Our model combines two main approaches based on biological features and heterogeneous networks to identify interactions between drugs and targets. First, we extract original features from the sequence to encode each drug and target. Later, we further consider the relationships among various biological entities by constructing drug-drug similarity networks and target-target similarity networks. Furthermore, a graph convolutional network and a graph attention network in the graph representation learning module help us learn the features representation of drugs and targets. After obtaining the features from graph representation learning modules, these features are combined into fusion descriptors for drug-target pairs. Finally, we send the fusion descriptors and labels to a random forest classifier for predicting DTI. The evaluation results show that BG-DTI achieves an average AUC of 0.938 and an average AUPR of 0.930, which is better than those of five existing state-of-the-art methods. We believe that BG-DTI can facilitate the development of drug discovery or drug repurposing.

Isolation, Screening, and Active Metabolites Identification of Anti-Vibrio Fungal Strains Derived From the Beibu Gulf Coral.

The Beibu Gulf harbors abundant underexplored marine microbial resources, which are rich in diversified secondary metabolites. The genera Vibrio is a well-known pathogenic bacterium of aquatic animals. In this study, 22 fungal strains were isolated and identified from the Beibu Gulf coral via the serial dilution method and internal transcribed spacer (ITS) sequence analysis, which were further divided into three branches by phylogenetic tree analysis. The crude extracts of them via small-scale fermentation were selected for the screening of inhibitory activity against Vibrio alginalyticus, Vibrio coralliilyticus, Vibrio harveyi, Vibrio parahaemolyticus, Vibrio owensii, and Vibrio shilonii. The results showed that eight fungal extracts displayed anti-Vibrio activity via the filter paper disk assay. Several of them showed strong inhibitory effects. Then, two tetramic acid alkaloids, equisetin (1) and 5'-epiequisetin (2), were identified from Fusarium equiseti BBG10 by bioassay-guided isolation, both of which inhibited the growth of Vibrio spp. with the MIC values of 86-132 µg/ml. The scanning electron microscope results showed that cell membranes of Vibrio became corrugated, distorted or ruptured after treatment with 1 and 2. Taken together, this study provided eight fungal isolates with anti-Vibrio potentials, and two alkaloid-type antibiotics were found with anti-Vibrio effects from the bioactive strain F. equiseti BBG10. Our findings highlight the importance of exploring promising microbes from the Beibu Gulf for the identification of anti-Vibrio for future antibiotic development.

Dual-modality optical coherence tomography and fluorescence tethered capsule endomicroscopy.

OCT tethered capsule endomicroscopy (TCE) is an emerging noninvasive diagnostic imaging technology for gastrointestinal (GI) tract disorders. OCT measures tissue reflectivity that provides morphologic image contrast, and thus is incapable of ascertaining molecular information that can be useful for improving diagnostic accuracy. Here, we introduce an extension to OCT TCE that includes a fluorescence (FL) imaging channel for attaining complementary, co-registered molecular contrast. We present the development of an OCT-FL TCE capsule and a portable, plug-and-play OCT-FL imaging system. The technology is validated in phantom experiments and feasibility is demonstrated in a methylene blue (MB)-stained swine esophageal injury model, ex vivo and in vivo.

A minus-end directed kinesin motor directs gravitropism in Physcomitrella patens.

Gravity is a critical environmental factor regulating directional growth and morphogenesis in plants, and gravitropism is the process by which plants perceive and respond to the gravity vector. The cytoskeleton is proposed to play important roles in gravitropism, but the underlying mechanisms are obscure. Here we use genetic screening in Physcomitrella patens, to identify a locus GTRC, that when mutated, reverses the direction of protonemal gravitropism. GTRC encodes a processive minus-end-directed KCHb kinesin, and its N-terminal, C-terminal and motor domains are all essential for transducing the gravity signal. Chimeric analysis between GTRC/KCHb and KCHa reveal a unique role for the N-terminus of GTRC in gravitropism. Further study shows that gravity-triggered normal asymmetric distribution of actin filaments in the tip of protonema is dependent on GTRC. Thus, our work identifies a microtubule-based cellular motor that determines the direction of plant gravitropism via mediating the asymmetric distribution of actin filaments.

Association of GRP78 promoter polymorphisms and serum GRP78 level with risk of asthenozoospermia.

PURPOSE: The aim of this study was undertaken to investigate the association of 78-kDa glucose-regulated protein (GRP78) gene promoter polymorphisms with risk of asthenozoospermia (AZS) men. In addition, we performed association analysis between GRP78 promoter mutations and serum GRP78 level in asthenozoospermia. METHODS: The study population comprised 400 subjects with AZS patients and 400 healthy controls. We assessed GRP78 rs3216733, rs17840761, and rs17840762 polymorphisms by using Snapshot SNP genotyping assays; serum GRP78 level was measured by enzyme-linked immunosorbent assay (ELISA). Semen quality was assessed by computer-assisted semen analysis. RESULTS: We found that rs3216733 was associated with increased risk of AZS (Gd vs. dd: adjusted OR = 1.42, 95% CI, 1.06-1.93, P = 0.020; Gd/GG vs. dd: adjusted OR = 1.43, 95% CI, 1.08-1.91, P = 0.013; G vs. d adjusted OR = 1.26, 95% CI, 1.03-1.56, P = 0.027). The haplotype analyses showed the frequency of G-C-C haplotype was significantly higher in AZS (P = 0.026). The percentage of progressive motility sperm was lower in the asthenozoospermic men with Gd and Gd/GG genotypes than dd genotype (P = 0.003). Moreover, the serum GRP78 levels were significantly lower in rs3216733 Gd/GG genotypes compared with the dd genotype (P < 0.001). CONCLUSION: Our findings suggest that rs3216733 Gd/GG genotypes contribute to poor sperm motility, probably by decreasing the level of GRP78.

Heterotrimeric kinesin-2 (KIF3) mediates transition zone and axoneme formation of mouse photoreceptors.

Anterograde intraflagellar transport (IFT) employing kinesin-2 molecular motors has been implicated in trafficking of photoreceptor outer segment proteins. We generated embryonic retina-specific (prefix "emb") and adult tamoxifen-induced (prefix "tam") deletions of KIF3a and IFT88 in adult mice to study photoreceptor ciliogenesis and protein trafficking. In (emb)Kif3a(-/-) and in (emb)Ift88(-/-) mice, basal bodies failed to extend transition zones (connecting cilia) with outer segments, and visual pigments mistrafficked. In contrast, (tam)Kif3a(-/-) and (tam)Ift88(-/-) photoreceptor axonemes disintegrated slowly post-induction, starting distally, but rhodopsin and cone pigments trafficked normally for more than 2 weeks, a time interval during which the outer segment is completely renewed. The results demonstrate that visual pigments transport to the retinal outer segment despite removal of KIF3 and IFT88, and KIF3-mediated anterograde IFT is responsible for photoreceptor transition zone and axoneme formation.