Molecular substructure tree generative model for de novo drug design.

Deep learning shortens the cycle of the drug discovery for its success in extracting features of molecules and proteins. Generating new molecules with deep learning methods could enlarge the molecule space and obtain molecules with specific properties. However, it is also a challenging task considering that the connections between atoms are constrained by chemical rules. Aiming at generating and optimizing new valid molecules, this article proposed Molecular Substructure Tree Generative Model, in which the molecule is generated by adding substructure gradually. The proposed model is based on the Variational Auto-Encoder architecture, which uses the encoder to map molecules to the latent vector space, and then builds an autoregressive generative model as a decoder to generate new molecules from Gaussian distribution. At the same time, for the molecular optimization task, a molecular optimization model based on CycleGAN was constructed. Experiments showed that the model could generate valid and novel molecules, and the optimized model effectively improves the molecular properties.

Structural and functional changes in the default mode network in drug-resistant epilepsy.

PURPOSE: To investigate brain network properties and connectivity abnormalities of the default mode network (DMN) in drug-resistant epilepsy (DRE). The study was based on probabilistic fiber tracking and functional connectivity (FC) analysis, to explore the structural and functional connectivity patterns change between frontal lobe epilepsy (FLE) and temporal lobe epilepsy (TLE). METHODS: A total of 33 DRE patients (18 TLE and 15 FLE) and 30 healthy controls (HCs) were recruited. The volume fraction of the septal brain region of the DMN in DRE was calculated using FreeSurfer. The FC analysis was performed using Data Processing and Analysis for Brain Imaging in MATLAB. The structural connections between brain regions of the DMN were calculated based on probabilistic fiber tracking. RESULTS: The left precuneus (PCUN) volumes in epilepsy groups were lower than that in HCs. Compared with FLE, TLE showed reduced FC between the left hippocampus (HIP) and PCUN/medial frontal gyrus, and between the right inferior parietal lobule (IPL) and right superior temporal gyrus. Compared with HCs, FLE showed increased FCs between the right IPL and occipital lobe, and between the left superior frontal gyrus (SFG) and bilateral superior temporal gyrus. In terms of structural connectivity, TLE exhibited increased connectivity strength between the left SFG and left PCUN, and showed reduced connection strength between the left HIP and left posterior cingulate gyrus/left PCUN, when compared with the FLE. CONCLUSIONS: TLE and FLE patients showed structural and functional changes in the DMN. Compared with FLE patients, the TLE patients showed reduced structural and functional connection strengths between the left HIP and PCUN. These alterations in connection strengths holds promise for the identification of TLE and FLE.

Sequence-based drug-target affinity prediction using weighted graph neural networks.

BACKGROUND: Affinity prediction between molecule and protein is an important step of virtual screening, which is usually called drug-target affinity (DTA) prediction. Its accuracy directly influences the progress of drug development. Sequence-based drug-target affinity prediction can predict the affinity according to protein sequence, which is fast and can be applied to large datasets. However, due to the lack of protein structure information, the accuracy needs to be improved. RESULTS: The proposed model which is called WGNN-DTA can be competent in drug-target affinity (DTA) and compound-protein interaction (CPI) prediction tasks. Various experiments are designed to verify the performance of the proposed method in different scenarios, which proves that WGNN-DTA has the advantages of simplicity and high accuracy. Moreover, because it does not need complex steps such as multiple sequence alignment (MSA), it has fast execution speed, and can be suitable for the screening of large databases. CONCLUSION: We construct protein and molecular graphs through sequence and SMILES that can effectively reflect their structures. To utilize the detail contact information of protein, graph neural network is used to extract features and predict the binding affinity based on the graphs, which is called weighted graph neural networks drug-target affinity predictor (WGNN-DTA). The proposed method has the advantages of simplicity and high accuracy.

Clinical features of NOTCH2NLC-related neuronal intranuclear inclusion disease.

BACKGROUND: Abnormal expanded GGC repeats within the NOTCH2HLC gene has been confirmed as the genetic mechanism for most Asian patients with neuronal intranuclear inclusion disease (NIID). This cross-sectional observational study aimed to characterise the clinical features of NOTCH2NLC-related NIID in China. METHODS: Patients with NOTCH2NLC-related NIID underwent an evaluation of clinical symptoms, a neuropsychological assessment, electrophysiological examination, MRI and skin biopsy. RESULTS: In the 247 patients with NOTCH2NLC-related NIID, 149 cases were sporadic, while 98 had a positive family history. The most common manifestations were paroxysmal symptoms (66.8%), autonomic dysfunction (64.0%), movement disorders (50.2%), cognitive impairment (49.4%) and muscle weakness (30.8%). Based on the initial presentation and main symptomology, NIID was divided into four subgroups: dementia dominant (n=94), movement disorder dominant (n=63), paroxysmal symptom dominant (n=61) and muscle weakness dominant (n=29). Clinical (42.7%) and subclinical (49.1%) peripheral neuropathies were common in all types. Typical diffusion-weighted imaging subcortical lace signs were more frequent in patients with dementia (93.9%) and paroxysmal symptoms types (94.9%) than in those with muscle weakness (50.0%) and movement disorders types (86.4%). GGC repeat sizes were negatively correlated with age of onset (r=-0.196, p<0.05), and in the muscle weakness-dominant type (median 155.00), the number of repeats was much higher than in the other three groups (p<0.05). In NIID pedigrees, significant genetic anticipation was observed (p<0.05) without repeat instability (p=0.454) during transmission. CONCLUSIONS: NIID is not rare; however, it is usually misdiagnosed as other diseases. Our results help to extend the known clinical spectrum of NOTCH2NLC-related NIID.

Enhancing Protein Function Prediction Performance by Utilizing AlphaFold-Predicted Protein Structures.

The structure of a protein is of great importance in determining its functionality, and this characteristic can be leveraged to train data-driven prediction models. However, the limited number of available protein structures severely limits the performance of these models. AlphaFold2 and its open-source data set of predicted protein structures have provided a promising solution to this problem, and these predicted structures are expected to benefit the model performance by increasing the number of training samples. In this work, we constructed a new data set that acted as a benchmark and implemented a state-of-the-art structure-based approach for determining whether the performance of the function prediction model can be improved by putting additional AlphaFold-predicted structures into the training set and further compared the performance differences between two models separately trained with real structures only and AlphaFold-predicted structures only. Experimental results indicated that structure-based protein function prediction models could benefit from virtual training data consisting of AlphaFold-predicted structures. First, model performances were improved in all three categories of Gene Ontology terms (GO terms) after adding predicted structures as training samples. Second, the model trained only on AlphaFold-predicted virtual samples achieved comparable performances to the model based on experimentally solved real structures, suggesting that predicted structures were almost equally effective in predicting protein functionality.

The long-term effects of microplastics on soil organomineral complexes and bacterial communities from controlled-release fertilizer residual coating.

Controlled-release fertilizer (CRF) was applied widely in China as an efficient utilization strategy for improving grain yield and reducing the nitrogen contamination. However, it was indeterminate to know the impacts of inevitably imported plastic into the soil on sustainable development. After ten-year fixed-site experiment, the visible residual coating microplastics were separated from the soil to measure their changes, then the long-term effects of CRF application (theoretical microplastic content 0.018-0.151 g kg-1 soil) on soil architecture and bacterial communities were evaluated. Based on soil organomineral complexes (OMC) distribution experiments and soil 16S rRNA sequence analysis, residual coating microplastics had no significant impact on soil architecture and limited effects on soil bacteria, but became the specific microbial habitat. The nitrogen rate and nitrogen release mode affected sand- and silt-grade OMC, and nitrogen rate impacted soil bacteria communities. The residual coating, small inert particles, is safe for soil OMC and bacterial communities in agricultural soil. Due to the effectiveness of CRF on reducing environmental pollution, CRF is considered as a favorable measure to the sustainable agricultural development in Shandong Province, China.

Alterations in white matter integrity and asymmetry in patients with benign childhood epilepsy with centrotemporal spikes and childhood absence epilepsy: An automated fiber quantification tractography study.

PURPOSE: To investigate whether patients with benign childhood epilepsy with centrotemporal spikes (BECTS) and childhood absence epilepsy (CAE) show distinct patterns of white matter (WM) alterations and structural asymmetry compared with healthy controls and the relationship between WM alterations and epilepsy-related clinical variables. METHODS: We used automated fiber quantification to create tract profiles of fractional anisotropy (FA) and mean diffusivity (MD) in twenty-six patients with BECTS, twenty-nine patients with CAE, and twenty-four healthy controls. Group differences in FA and MD were quantified at 100 equidistant nodes along the fiber tract and these alterations and epilepsy-related clinical variables were correlated. A lateralization index (LI) representing the structural asymmetry of the fiber tract was computed and compared between both patient groups and controls. RESULTS: Compared with healthy controls, the BECTS group showed widespread FA reduction in 43.75% (7/16) and MD elevation in 50% (8/16) of identified fiber tracts, and the CAE group showed regional FA reduction in 31.25% (5/16) and MD elevation in 25% (4/16) of identified fiber tracts. In the BECTS group, FA and MD in the right anterior thalamic radiation positively and negatively correlated with the number of antiepileptic drugs, respectively, and MD in the right arcuate fasciculus (AF) positively correlated with seizure frequency. In the CAE group, the LI values were significantly lower in the inferior fronto-occipital fasciculus and the AF. CONCLUSION: The two childhood epilepsy syndromes display different patterns of WM alterations and structural asymmetry, suggesting that neuroanatomical differences may underlie the different profiles of BECTS and CAE.

SAG-DTA: Prediction of Drug-Target Affinity Using Self-Attention Graph Network.

The prediction of drug-target affinity (DTA) is a crucial step for drug screening and discovery. In this study, a new graph-based prediction model named SAG-DTA (self-attention graph drug-target affinity) was implemented. Unlike previous graph-based methods, the proposed model utilized self-attention mechanisms on the drug molecular graph to obtain effective representations of drugs for DTA prediction. Features of each atom node in the molecular graph were weighted using an attention score before being aggregated as molecule representation. Various self-attention scoring methods were compared in this study. In addition, two pooing architectures, namely, global and hierarchical architectures, were presented and evaluated on benchmark datasets. Results of comparative experiments on both regression and binary classification tasks showed that SAG-DTA was superior to previous sequence-based or other graph-based methods and exhibited good generalization ability.

Molecule Property Prediction Based on Spatial Graph Embedding.

Accurate prediction of molecular properties is important for new compound design, which is a crucial step in drug discovery. In this paper, molecular graph data is utilized for property prediction based on graph convolution neural networks. In addition, a convolution spatial graph embedding layer (C-SGEL) is introduced to retain the spatial connection information on molecules. And, multiple C-SGELs are stacked to construct a convolution spatial graph embedding network (C-SGEN) for end-to-end representation learning. In order to enhance the robustness of the network, molecular fingerprints are also combined with C-SGEN to build a composite model for predicting molecular properties. Our comparative experiments have shown that our method is accurate and achieves the best results on some open benchmark datasets.

Toxicity Prediction Method Based on Multi-Channel Convolutional Neural Network.

Molecular toxicity prediction is one of the key studies in drug design. In this paper, a deep learning network based on a two-dimension grid of molecules is proposed to predict toxicity. At first, the van der Waals force and hydrogen bond were calculated according to different descriptors of molecules, and multi-channel grids were generated, which could discover more detail and helpful molecular information for toxicity prediction. The generated grids were fed into a convolutional neural network to obtain the result. A Tox21 dataset was used for the evaluation. This dataset contains more than 12,000 molecules. It can be seen from the experiment that the proposed method performs better compared to other traditional deep learning and machine learning methods.

Controlled-release urea reduced nitrogen leaching and improved nitrogen use efficiency and yield of direct-seeded rice.

The use of controlled-release urea (CRU) has become one of best management practices for increasing crop yield and improving nitrogen (N) use efficiency (NUE). However, the effects of CRU on direct-seeded rice are not well understood while direct-seeding has gradually replaced transplanting due to increasing labor cost and lack of irrigation water. The objective of this two-year field experiment was to compare the effects of the CRU at four rates (120, 180, 240 and 360 kg N ha-1, CRU1, CRU2, CRU3 and CRU4, respectively) with a conventional urea fertilizer (360 kg N ha-1; U) and a control (no N fertilizer applied; CK) on yield, biomass, NUE of direct-seeded rice and soil nutrients. The results indicated that the successive release rates of N from CRU corresponded well to the N requirements of rice. The use of CRU3 and CRU4 increased rice grain yields by 20.8 and 28.7%, respectively, compared with U. In addition, the NUEs were improved by all CRU treatments compared to the U treatment. Concentrations of NO3--N and NH4+-N in the soil were increased, especially during the later growth stages of the rice, and the leaching of N was reduced with CRU treatments. In conclusion, applying CRU on direct-seeded rice increased the crops yields and NUE, increased nitrogen availability at the late growth stages, and reduced N leaching.

Hierarchical Activity Recognition Using Smart Watches and RGB-Depth Cameras.

Human activity recognition is important for healthcare and lifestyle evaluation. In this paper, a novel method for activity recognition by jointly considering motion sensor data recorded by wearable smart watches and image data captured by RGB-Depth (RGB-D) cameras is presented. A normalized cross correlation based mapping method is implemented to establish association between motion sensor data with corresponding image data from the same person in multi-person situations. Further, to improve the performance and accuracy of recognition, a hierarchical structure embedded with an automatic group selection method is proposed. Through this method, if the number of activities to be classified is changed, the structure will be changed correspondingly without interaction. Our comparative experiments against the single data source and single layer methods have shown that our method is more accurate and robust.

Computational analysis of long QT syndrome type 2 and the therapeutic effects of KCNQ1 antibodies.

Objective: Long QT interval syndrome (LQTS) is a highly dangerous cardiac disease that can lead to sudden cardiac death; however, its underlying mechanism remains largely unknown. This study is conceived to investigate the impact of two general genotypes of LQTS type 2, and also the therapeutic effects of an emerging immunology-based treatment named KCNQ1 antibody. Methods: A multiscale virtual heart is developed, which contains multiple biological levels ranging from ion channels to a three-dimensional cardiac structure with realistic geometry. Critical biomarkers at different biological levels are monitored to investigate the remodeling of cardiac electrophysiology induced by mutations. Results: Simulations revealed multiple important mechanisms that are hard to capture via conventional clinical techniques, including the augmented dispersion of repolarization, the increased vulnerability to arrhythmias, the impaired adaptability in tissue to high heart rates, and so on. An emerging KCNQ1 antibody-based therapy could rescue the prolonged QT interval but did not reduce the vulnerable window. Conclusions: Tiny molecular alterations can lead to cardiac electrophysiological remodeling at multiple biological levels, which in turn contributes to higher susceptibility to lethal arrhythmias in long QT syndrome type 2 patients. The KCNQ1 antibody-based therapy has proarrhythmic risks notwithstanding its QT-rescuing effects.

CollaPPI: A Collaborative Learning Framework for Predicting Protein-Protein Interactions.

Exploring protein-protein interaction (PPI) is of paramount importance for elucidating the intrinsic mechanism of various biological processes. Nevertheless, experimental determination of PPI can be both time-consuming and expensive, motivating the exploration of data-driven deep learning technologies as a viable, efficient, and accurate alternative. Nonetheless, most current deep learning-based methods regarded a pair of proteins to be predicted for possible interaction as two separate entities when extracting PPI features, thus neglecting the knowledge sharing among the collaborative protein and the target protein. Aiming at the above issue, a collaborative learning framework CollaPPI was proposed in this study, where two kinds of collaboration, i.e., protein-level collaboration and task-level collaboration, were incorporated to achieve not only the knowledge-sharing between a pair of proteins, but also the complementation of such shared knowledge between biological domains closely related to PPI (i.e., protein function, and subcellular location). Evaluation results demonstrated that CollaPPI obtained superior performance compared to state-of-the-art methods on two PPI benchmarks. Besides, evaluation results of CollaPPI on the additional PPI type prediction task further proved its excellent generalization ability.

Optimization of Synthesis Conditions for Urea-Formaldehyde Slow-Release Fertilizer Using Response Surface Methodology.

In this study, we optimized the preparation of urea-formaldehyde fertilizer using response surface methodology with a Box-Behnken experimental design. The aim was to maximize the difference between CWIR and HWIR to maximize the content of slow-release insoluble nitrogen. In this work, a model of the impact of reaction factors on CWIR and HWIR was established. Through analysis of variance, the final model was significant. According to this model, the optimal reaction conditions were: a reaction temperature of 42.5 °C, a reaction time of 66.2 min, a U/F of 1.68, and a pH 3.3. Under these optimal conditions, the CWIR and HWIR reached 55.65 and 33.92%, respectively. In addition, the samples were characterized by scanning electron microscopy and thermal stability analysis. This study accurately synthesized urea-formaldehyde products with specific release periods according to production needs in order to improve the efficiency of fertilizer utilization.

Preparation and characterization of liquefied eggplant branch bio-based controlled-release fertilizer.

Bio-based coating materials have received increased attention because of their low-cost, environmentally friendly, and sustainable properties. In this paper, a novel coating material was developed to coat ureas using bio-based coating material derived from liquefied eggplant branches to form controlled-release ureas (CRUs). Also, the optimum proportion of liquefier was studied. Furthermore, dimethyl siloxane was used to modify liquified eggplant branches to make them hydrophobic, resulting in hydrophobic controlled-release ureas (SCRUs). This hydrophobic-enabled coating is environmentally friendly and highly efficient. The products were characterized by specific scanning electron microscopy, energy-dispersive X-ray spectroscopy, Fourier transform infrared spectroscopy, thermogravimetric analysis, and differential scanning calorimetry, and the water contact angles of CRUs and SCRUs were determined. The nutrient-release characteristics of the SCRUs in water were determined at 25 °C and compared with those of CRUs. The results showed that the modification with dimethyl siloxane reduced the N release rate and increased the longevity of the fertilizer coated with hydrophobic bio-based coating material. In addition, organosilicon atoms on the SCRU surface also block the micro-holes on the coating and thus reduce the entry of water onto the coating. The results suggest that the new coating technology can create a hydrophobic surface on bio-based coating material and thus improve their controlled-release characteristics.

A sustainable lignin-clay nano-absorbent for highly efficient methylene blue removal.

Water pollution caused by toxic dyes such as methylene blue (MB) has become a bottleneck for recycling or reusing enormous industrial wastewater. Designing a green and cost-effective bio-absorbent for the highly efficient removal of MB from wastewater is crucial but remains a great challenge. In this study, abundant, inexpensive, and environmentally benign lignin and bentonite were used as starting materials, and quaternary and amphiphilic lignin as a network macromolecule was designed to be inserted into the galleries of the stacked bentonite clay to prepare lignin-bentonite nanohybrids. The specific surface area of the modified nano-absorbent was significantly increased to 45.602 m2/g and owned a type II-like isothermal mechanism. The absorbent showed a maximum removal of MB of 99.7 % at neutral pH and room temperature with a maximum adsorption capacity of 822.22 mg/g, demonstrating the potential of an excellent adsorbent. The MB adsorption process fits well with both Langmuir and Freundlich isotherm models, the adsorption mechanism includes strong electrostatic attraction between absorbent and MB, and physical adsorption in a complicated monolayer and multiple macroporous structures. This study leverages the economic and environmental benefits of lignin and bentonite clay to prepare a cost-effective bio-absorbent for efficient removal of MB from aqueous solutions.

Complexes of glycated casein and carboxymethyl cellulose enhance stability and control release of anthocyanins.

To improve the stability and sustained-release property of anthocyanins (ACNs), casein (CA) - dextran (DEX) glycated conjugates (UGCA) and carboxymethyl cellulose (CMC) were used to prepare ACNs-loaded binary and ternary complexes. The ACNs-loaded binary complexes (ACNs-UGCA) and ternary complexes (ACNs-UGCA-CMC) achieved by 8 min' ultrasonic treatment with 40 % amplitude. The binary and ternary complexes showed spherical structure and good dispersibility, with the average size of 121.2 nm and 132.4 nm respectively. The anthocyanins encapsulation efficiency of ACNs-UGCA-CMC increased almost 20 % than ACNs-UGCA. ACNs-UGCA-CMC had better colloidal stabilities than ACNs-UGCA, such as thermal stability and dilution stability. Simultaneously, both of the binary and ternary complexes significantly prevented anthocyanins from being degraded by heat treatment, ascorbic acid, sucrose and simulated gastrointestinal environment. The protective effect of ACNs-UGCA-CMC was more significant. Furthermore, ACNs-UGCA-CMC showed slower anthocyanins release in simulated releasing environment in vitro and a long retention time in vivo. Our current study provides a potential delivery for improving the stability and controlling release of anthocyanins.

HiSIF-DTA: A Hierarchical Semantic Information Fusion Framework for Drug-Target Affinity Prediction.

Accurately identifying drug-target affinity (DTA) plays a significant role in promoting drug discovery and has attracted increasing attention in recent years. Exploring appropriate protein representation methods and increasing the abundance of protein information is critical in enhancing the accuracy of DTA prediction. Recently, numerous deep learning-based models have been proposed to utilize the sequential or structural features of target proteins. However, these models capture only the low-order semantics that exist in a single protein, while the high-order semantics abundant in biological networks are largely ignored. In this article, we propose HiSIF-DTA'a hierarchical semantic information fusion framework for DTA prediction. In this framework, a hierarchical protein graph is constructed that includes not only contact maps as low-order structural semantics but also protein-rotein interaction (PPI) networks as high-order functional semantics. Particularly, two distinct hierarchical fusion strategies (i.e., Top-down and Bottom-Up) are designed to integrate the different protein semantics, therefore contributing to a richer protein representation. Comprehensive experimental results demonstrate that HiSIF-DTA outperforms current state-of-the-art methods for prediction on the benchmark datasets of the DTA task. Further validation on binary tasks and visualization analysis demonstrates the generalization and interpretation abilities of the proposed method.

Structural and functional differences of the thalamus between drug-naïve Parkinson's disease motor subtypes.

Objective: The thalamus is an integrative hub of motor circuits in Parkinson's disease (PD). This study aimed to investigate the alterations of structure and functional connectivity (FC) of the thalamic subregions in the tremor-dominant (TD) subtype and the postural instability and gait difficulty (PIGD) subtype in PD. Methods: A total of 59 drug-naïve patients (24 TD and 35 PIGD) and 37 healthy controls were recruited. The volumes of the thalamus and the thalamic subregions were calculated using FreeSurfer. Functional connectivity (FC) analysis of the resting-state functional MRI (rsfMRI) was conducted on the thalamic subregions. Finally, the altered structure and FC were used for correlation analysis with clinical motor scores and for further motor subtypes differentiation. Results: The volumes of the left posterior parietal thalamus (PPtha) in TD patients were significantly lower than those of PIGD patients. Compared with PIGD patients, TD patients exhibited higher FC between the thalamic subregions, the left middle temporal gyrus (MTG), the right dorsolateral superior frontal gyrus (SFGdl), the left middle occipital gyrus (MOG), and the right superior temporal gyrus (STG). Compared with HCs, TD patients showed higher FC between the thalamic subregions and the right SFGdl, as well as the left MOG. Compared with HCs, PIGD patients showed lower FC between the thalamic subregions and the left MTG. In addition, the altered FC was closely related to clinical symptoms and performed high-discriminative power in differentiating the motor subtypes. Conclusion: Increased FC between the thalamic subregions and the sensory cortices in TD patients may indicate a better compensatory capacity for impairment of sensory information integration than that in PIGD patients. The altered FC between the thalamus and the MTG was a potential biomarker for the distinction of the PD motor subtypes.