AMGDTI: drug-target interaction prediction based on adaptive meta-graph learning in heterogeneous network.

Prediction of drug-target interactions (DTIs) is essential in medicine field, since it benefits the identification of molecular structures potentially interacting with drugs and facilitates the discovery and reposition of drugs. Recently, much attention has been attracted to network representation learning to learn rich information from heterogeneous data. Although network representation learning algorithms have achieved success in predicting DTI, several manually designed meta-graphs limit the capability of extracting complex semantic information. To address the problem, we introduce an adaptive meta-graph-based method, termed AMGDTI, for DTI prediction. In the proposed AMGDTI, the semantic information is automatically aggregated from a heterogeneous network by training an adaptive meta-graph, thereby achieving efficient information integration without requiring domain knowledge. The effectiveness of the proposed AMGDTI is verified on two benchmark datasets. Experimental results demonstrate that the AMGDTI method overall outperforms eight state-of-the-art methods in predicting DTI and achieves the accurate identification of novel DTIs. It is also verified that the adaptive meta-graph exhibits flexibility and effectively captures complex fine-grained semantic information, enabling the learning of intricate heterogeneous network topology and the inference of potential drug-target relationship.

Comprehensive evaluation of deep and graph learning on drug-drug interactions prediction.

Recent advances and achievements of artificial intelligence (AI) as well as deep and graph learning models have established their usefulness in biomedical applications, especially in drug-drug interactions (DDIs). DDIs refer to a change in the effect of one drug to the presence of another drug in the human body, which plays an essential role in drug discovery and clinical research. DDIs prediction through traditional clinical trials and experiments is an expensive and time-consuming process. To correctly apply the advanced AI and deep learning, the developer and user meet various challenges such as the availability and encoding of data resources, and the design of computational methods. This review summarizes chemical structure based, network based, natural language processing based and hybrid methods, providing an updated and accessible guide to the broad researchers and development community with different domain knowledge. We introduce widely used molecular representation and describe the theoretical frameworks of graph neural network models for representing molecular structures. We present the advantages and disadvantages of deep and graph learning methods by performing comparative experiments. We discuss the potential technical challenges and highlight future directions of deep and graph learning models for accelerating DDIs prediction.

Combined associations of visceral adipose tissue and adherence to a Mediterranean lifestyle with T2D and diabetic microvascular complications among individuals with prediabetes.

BACKGROUND: It's unclear if excess visceral adipose tissue (VAT) mass in individuals with prediabetes can be countered by adherence to a Mediterranean lifestyle (MEDLIFE). We aimed to examine VAT mass, MEDLIFE adherence, and their impact on type 2 diabetes (T2D) and diabetic microvascular complications (DMC) in individuals with prediabetes. METHODS: 11,267 individuals with prediabetes from the UK Biobank cohort were included. VAT mass was predicted using a non-linear model, and adherence to the MEDLIFE was evaluated using the 25-item MEDLIFE index, encompassing categories such as "Mediterranean food consumption," "Mediterranean dietary habits," and "Physical activity, rest, social habits, and conviviality." Both VAT and MEDLIFE were categorized into quartiles, resulting in 16 combinations. Incident cases of T2D and related DMC were identified through clinical records. Cox proportional-hazards regression models were employed to examine associations, adjusting for potential confounding factors. RESULTS: Over a median follow-up of 13.77 years, we observed 1408 incident cases of T2D and 714 cases of any DMC. High adherence to the MEDLIFE, compared to the lowest quartile, reduced a 16% risk of incident T2D (HR: 0.84, 95% CI: 0.71-0.98) and 31% for incident DMC (0.69, 0.56-0.86). Conversely, compared to the lowest quartile of VAT, the highest quartile increased the risk of T2D (5.95, 4.72-7.49) and incident any DMC (1.79, 1.36-2.35). We observed an inverse dose-response relationship between MEDLIFE and T2D/DMC, and a dose-response relationship between VAT and all outcomes (P for trend < 0.05). Restricted cubic spline analysis confirmed a nearly linear dose-response pattern across all associations. Compared to individuals with the lowest MEDLIFE quartile and highest VAT quartile, those with the lowest T2D risk had the lowest VAT and highest MEDLIFE (0.12, 0.08-0.19). High MEDLIFE was linked to reduced T2D risk across all VAT categories, except in those with the highest VAT quartile. Similar trends were seen for DMC. CONCLUSION: High adherence to MEDLIFE reduced T2D and MDC risk in individuals with prediabetes, while high VAT mass increases it, but MEDLIFE adherence may offset VAT's risk partly. The Mediterranean lifestyle's adaptability to diverse populations suggests promise for preventing T2D.

A Discrete, Boron-Containing Triangular Triradical.

A neutral boron-containing triangular triradical based on a triptycene derivative has been designed and synthesized. Its structure, bonding and physical property have been studied by EPR spectroscopy, SQUID magnetometry and single crystal X-ray diffraction, as well as theoretical calculations. The triradical has a series of isosceles triangle conformations in the solution due to the Jahn-Teller distortion, leading to the splitting of the two low-lying doublet states. This factor together with negligible spin-orbit coupling (SOC) of composing light atoms quenches the spin frustration. The work represents a rare example of a neutral through-space triangular triradical.

A muti-modal feature fusion method based on deep learning for predicting immunotherapy response.

Immune checkpoint therapy (ICT) has greatly improved the survival of cancer patients in the past few years, but only a small number of patients respond to ICT. To predict ICT response, we developed a multi-modal feature fusion model based on deep learning (MFMDL). This model utilizes graph neural networks to map gene-gene relationships in gene networks to low dimensional vector spaces, and then fuses biological pathway features and immune cell infiltration features to make robust predictions of ICT. We used five datasets to validate the predictive performance of the MFMDL. These five datasets span multiple types of cancer, including melanoma, lung cancer, and gastric cancer. We found that the prediction performance of multi-modal feature fusion model based on deep learning is superior to other traditional ICT biomarkers, such as ICT targets or tumor microenvironment-associated markers. In addition, we also conducted ablation experiments to demonstrate the necessity of fusing different modal features, which can improve the prediction accuracy of the model.

Prediction of miRNA-disease associations by neural network-based deep matrix factorization.

MicroRNA(miRNA) is a class of short non-coding RNAs with a length of about 22 nucleotides, which participates in various biological processes of cells. A number of studies have shown that miRNAs are closely related to the occurrence of cancer and various human diseases. Therefore, studying miRNA-disease associations is helpful to understand the pathogenesis of diseases as well as the prevention, diagnosis, treatment and prognosis of diseases. Traditional biological experimental methods for studying miRNA-disease associations have disadvantages such as expensive equipment, time-consuming and labor-intensive. With the rapid development of bioinformatics, more and more researchers are committed to developing effective computational methods to predict miRNA-disease associations in roder to reduce the time and money cost of experiments. In this study, we proposed a neural network-based deep matrix factorization method named NNDMF to predict miRNA-disease associations. To address the problem that traditional matrix factorization methods can only extract linear features, NNDMF used neural network to perform deep matrix factorization to extract nonlinear features, which makes up for the shortcomings of traditional matrix factorization methods. We compared NNDMF with four previous classical prediction models (IMCMDA, GRMDA, SACMDA and ICFMDA) in global LOOCV and local LOOCV, respectively. The AUCs achieved by NNDMF in two cross-validation methods were 0.9340 and 0.8763, respectively. Furthermore, we conducted case studies on three important human diseases (lymphoma, colorectal cancer and lung cancer) to validate the effectiveness of NNDMF. In conclusion, NNDMF could effectively predict the potential miRNA-disease associations.

Differentiating Oxygen Exchange Reaction Mechanisms across Phase Boundaries.

Triggering phase transitions by controlling the anion stoichiometry is an effective method of tuning the electrocatalytic activity of the functional oxides. However, understanding the potential differences in the reaction mechanism(s) of different phases requires the accurate mapping of phase boundaries during the electrochemical reactions, which can be quite challenging. In this work, we have established a feasible electrochemical method based on the measurement of chemical capacitance to resolve the critical stoichiometry at phase boundaries under operando conditions. We select a simple binary oxide PrOx as a proof-of-principle model system, which shows excellent activity for high-temperature oxygen incorporation and evolution reactions (OIR/OER). We show that the phase transition can be sensitively probed by quantifying the chemical capacitance, which can be further used for differentiating the OIR/OER mechanisms across the phase boundary of PrOx. Therefore, our findings provide a new framework for exploring phase engineering as a tool for the design of electrocatalysts.

The independent and joint association of accelerometer-measured physical activity and sedentary time with dementia: a cohort study in the UK Biobank.

BACKGROUND: Research on the association of physical activity and sedentary time with dementia is accumulating, though elusive, and the interaction effects of the two remain unclear. We analysed the joint associations of accelerometer-measured physical activity and sedentary time with risk of incident dementia (all-cause dementia, Alzheimer's disease and vascular dementia). METHODS: A total of 90,320 individuals from the UK Biobank were included. Accelerometer-measured total volume of physical activity (TPA) and sedentary time were measured at baseline and dichotomised by median (low TPA [< 27 milli-gravity (milli-g)], high TPA [≥ 27 milli-g]; low sedentary time [< 10.7 h/day], high sedentary time [≥ 10.7 h/day]). Cox proportional hazards models were used to evaluate the joint associations with incident dementia on both additive and multiplicative scales. RESULTS: During a median follow-up of 6.9 years, 501 cases of all-cause dementia were identified. Higher TPA was associated with a lower risk of all-cause dementia, Alzheimer's disease and vascular dementia; the multivariate adjusted hazard ratios (HRs) (95% CI) per 10 milli-g increase were 0.63 (0.55-0.71), 0.74 (0.60-0.90) and 0.69 (0.51-0.93), respectively. Sedentary time was only found to be linked to all-cause dementia, and the HR for high sedentary time was 1.03 (1.01-1.06) compared with that for low sedentary time. No additive and multiplicative relationship of TPA and sedentary time to incident dementia was found (all P values > 0.05). CONCLUSION: Higher TPA level was related to a lower risk of incident dementia irrespective of sedentary time, which highlighted the implication of promoting physical activity participation to counteract the potential detrimental effect of sedentary time on dementia.

Dynamically tunable terahertz quadruple-function absorber based on a hybrid configuration of graphene and vanadium dioxide.

A quadruple-function dynamically tunable terahertz absorber that uses a hybrid configuration of graphene and vanadium dioxide is proposed in this paper. The absorber achieves dynamic conversion of four functions in one structure: ultra-broadband, broadband, single-frequency narrowband and dual-frequency narrowband, by utilizing the electrical control properties of graphene and the phase-shifting properties of vanadium dioxide. Furthermore, the paper also reveals the physical mechanism of the proposed absorber through the electric field distribution and impedance matching theory. In addition, the influences of the Fermi energy level of graphene and the electrical conductivity of vanadium dioxide on the absorption spectra are investigated, demonstrating the structure's dynamic tunability. Due to the above features, the designed absorber is expected to have potential applications in terahertz imaging, modulation and filtering.

Association of serum uric acid and fasting plasma glucose with cognitive function: a cross-sectional study.

BACKGROUND: The combined effect of serum uric acid (SUA) and blood glucose on cognition has not been explored. This study aimed to examine the separate and combined association of SUA and fasting plasma glucose (FPG) or diabetes mellitus (DM) with cognition in a sample of Chinese middle-aged and elderly population. METHODS: A total of 6,509 participants aged 45 years or older who participated in the China Health and Retirement Longitudinal Study (CHARLS, 2011) were included. The three cognitive domains assessed were episodic memory, mental status, and global cognition (the sum of the first two terms). Higher scores indicated better cognition. SUA and FPG were measured. The participants were grouped based on SUA and FPG quartiles to evaluate their combined associations of cognition with SUA Q1-Q3 only (Low SUA), with FPG Q4 only (High FPG), without low SUA and high FPG levels (Non), and with low SUA and high FPG levels (Both), multivariate linear regression models were used to analyze their association. RESULTS: Lower SUA quartiles were associated with poorer performance in global cognition and episodic memory compared with the highest quartile. Although no association was found between FPG or DM and cognition, high FPG or DM combined with low SUA levels in women (ßFPG = -0.983, 95% CI: -1.563--0.402; ßDM = -0.800, 95% CI: -1.369--0.232) had poorer cognition than those with low SUA level only (ßFPG = -0.469, 95% CI: -0.926--0.013; ßDM = -0.667, 95% CI: -1.060--0.275). CONCLUSION: Maintaining an appropriate level of SUA may be important to prevent cognitive impairment in women with high FPG.

Protonation-Induced Colossal Chemical Expansion and Property Tuning in NdNiO3 Revealed by Proton Concentration Gradient Thin Films.

Protonation can be used to tune diverse physical and chemical properties of functional oxides. Although protonation of nickelate perovskites has been reported, details on the crystal structure of the protonated phase and a quantitative understanding of the effect of protons on physical properties are still lacking. Therefore, in this work, we select NdNiO3 (NNO) as a model system to understand the protonation process from pristine NNO to protonated HxNdNiO3 (H-NNO). We used a reliable electrochemical method with well-defined reference electrode to trigger the protonation-induced phase transition. We found that the protonated H-NNO phase showed a colossal ∼13% lattice expansion caused by a large tilt of NiO6 octahedra and displacement of Nd cations. Importantly, we further designed a novel device configuration to induce a gradient of proton concentration into a single NNO thin film to establish a quantitative correlation between the proton concentration and the lattice constant and transport property of H-NNO.

A Simple Molecular Design Strategy for Pure-Red Multiple Resonance Emitters.

Though the flourishment of materials with multiple resonance (MR) in blue to green regions, red-emissive MR emitters are still rare in literatures, which definitely should be resolved for further applications. Herein, we report a simple molecular design strategy for the construction of pure-red MR emitters by conjugate charge transfer, which could greatly enhance the π-conjugation degree and charge-transfer property of the target molecule while maintaining the basic feature of MR, leading to a significant redshift of more than 128 nm compared to the selected parent MR core. The proof-of-concept emitter PPZ-BN exhibited a pure-red emission with a dominant peak at 613 nm and a small full-width-at-half-maximum of 0.16 eV (48 nm). The optimized organic light-emitting diode showed a high external quantum efficiency of 26.9 %, a small efficiency roll-off, and an excellent operation stability (LT99) of more than 43 hours at an initial luminance of 10 000 cd m-2 .

Exceeding 30 % External Quantum Efficiency in Non-doped OLEDs Utilizing Solution Processable TADF Emitters with High Horizontal Dipole Orientation via Anchoring Strategy.

Strategies to enhance the ratio of the molecular horizontal emitting dipole orientation (Θ∥ ) for thermally activated delayed fluorescence (TADF) emitters have unlocked the full potential of efficiencies for the evaporated devices, which, however, remain elusive for the solution-processed ones. Here, a strategic molecular design for solution processable TADF emitters featuring high Θ∥ s is proposed by attaching flexible chains ended with bipolar 9,9'-spirobi[fluorene] subunits as anchoring groups onto TADF emitting core. It's unveiled that the anchoring groups not only enhance the horizontal orientation via enlarging molecular planarity, but also benefit the high photoluminescence in pristine films. The corresponding non-doped solution processable OLEDs substantiate an unprecedented maximum external quantum efficiency (EQEmax )>30 %. Meanwhile, combining these compounds as TADF sensitizers, and multiple resonance final emitter, solution-processed OLEDs achieve an EQEmax of 25.6 % with a narrow full width at half maximum of 29 nm.

The incidence of mental disorder increases after hip fracture in older people: a nationwide cohort study.

BACKGROUND: People living with dementia seem to be more likely to experience delirium following hip fracture. The association between mental disorders (MD) and hip fracture remains controversial. We conducted a nationwide study to examine the prevalence of MD in geriatric patients with hip fractures undergoing surgery and conducted a related risk factor analysis. MATERIAL AND METHODS: This retrospective cohort study used data from Taiwan's National Health Insurance Research Database between 2000 and 2012 and focused on people who were older than 60 years. Patients with hip fracture undergoing surgical intervention and without hip fracture were matched at a ratio of 1:1 for age, sex, comorbidities, and index year. The incidence and hazard ratios of age, sex, and multiple comorbidities related to MD and its subgroups were calculated using Cox proportional hazards regression models. RESULTS: A total of 1408 patients in the hip fracture group and a total of 1408 patients in the control group (no fracture) were included. The overall incidence of MD for the hip fracture and control groups per 100 person-years were 0.8 and 0.5, respectively. Among MD, the incidences of transient MD, depression, and dementia were significantly higher in the hip fracture group than in the control group. CONCLUSIONS: The prevalence of newly developed MD, especially transient MD, depression, and dementia, was higher in the geriatric patients with hip fracture undergoing surgery than that in the control group. Prompt and aggressive prevention protocols and persistent follow-up of MD development is highly necessary in this aged society.

Hepatic encephalopathy increases the risk of hip fracture: a nationwide cohort study.

BACKGROUND: Osteoporotic hip fracture is a common general health problem with a significant impact on human life because it debilitates the patients and largely decreases their quality of life. Early prevention of fractures has become essential in recent decades. This can be achieved by evaluating the related risk factors, as a reference for further intervention. This is especially useful for the vulnerable patient group with comorbidities. Hepatic encephalopathy (HE), a major complication of liver cirrhosis, may increase the rate of falls and weaken the bone. This study evaluated the correlation between hepatic encephalopathy and osteoporotic hip fracture in the aged population using a national database. METHODS: This retrospective cohort study used data from Taiwan's National Health Insurance Research Database between 2000 and 2012. We included people who were older than 50 years with hepatic encephalopathy or other common chronic illnesses. Patients with and without hepatic encephalopathy were matched at a ratio of 1:4 for age, sex, and index year. The incidence and hazard ratios of osteoporotic hip fracture between the both cohorts were calculated using Cox proportional hazard regression models. RESULTS: The mean age of the enrolled patients was 66.5 years. The incidence ratio of osteoporotic hip fracture in the HE group was significantly higher than that in the non-HE group (68/2496 [2.7%] vs 98/9984 [0.98%]). Patients with HE were 2.15-times more likely to develop osteoporotic hip fractures than patients without HE in the whole group. The risk ratio was also significantly higher in female and older individuals. The results were also similar in the comorbidity subgroups of hypertension, diabetes mellitus, hyperlipidemia, senile cataract, gastric ulcer, and depression. Alcohol-related illnesses seemed to not confound the results of this study. CONCLUSIONS: HE is significantly associated with an increased risk of osteoporotic hip fractures, and the significance is not affected by the comorbidities in people aged more than 50 years. The cumulative risk of fracture increases with age.

Influence of two-photon absorption on the dynamic behaviors of microring resonators.

In this paper, we have investigated the influence of two-photon absorption (TPA) on the dynamic behaviors of all-pass and add-drop microring resonators by using two iterative methods along with the linear stability analysis method. While the incident field is above a certain value, the TPA coefficient has greater influence on the steady state for all-pass and add-drop microring resonators. We use the linear stability analysis method to analyze the stability of the steady state solutions and obtain stability conditions. Results obtained have shown that the change of TPA coefficient will lead to different dynamic behaviors; in addition, while the TPA coefficient is small and its change is slight, the dynamic behaviors of the microring resonators will not change much for most regions. At last, we observe the period windows and route from chaotic to period-N in some original chaotic regions due to the fluctuation of the TPA coefficient.

A new technique for generating pathogenic barcodes in breast cancer susceptibility analysis.

Complex diseases usually involve complex interactions between multiple loci. The artificial intelligent algorithm is a plausible strategy to evade combinatorial explosion. However, the randomness of solution of this algorithm loses decreases the confidence of biological researchers on this algorithm. Meanwhile, the lack of an efficient and effective measure to profile the distribution of cases and controls impedes the discovery of pathogenic epistasis. Here we present an efficient method called maximum dissimilarity-minimum entropy (MDME) to analyze breast cancer single-nucleotide polymorphism (SNP) data. The method searches risky barcodes, which to increase the odds ratio and relative risk of the breast cancer. This method based on the hypothesis that if a specific barcode is associated with a disease, then the barcode permits distinction of cases from controls and more importantly it shows a relative consistent pattern in cases. An analysis based on simulated dataset explains the necessity of minimum entropy. Experimental results show that our method can find the most risky barcode that contributes to breast cancer susceptibility. Our method may also mine several pathogenic barcodes that condition the different subtypes of cancer.

Fourier-sparsity integrated method for complex target ISAR imagery.

In existing sparsity-driven inverse synthetic aperture radar (ISAR) imaging framework a sparse recovery (SR) algorithm is usually applied to azimuth compression to achieve high resolution in the cross-range direction. For range compression, however, direct application of an SR algorithm is not very effective because the scattering centers resolved in the high resolution range profiles at different view angles always exhibit irregular range cell migration (RCM), especially for complex targets, which will blur the ISAR image. To alleviate the sparse recovery-induced RCM in range compression, a sparsity-driven framework for ISAR imaging named Fourier-sparsity integrated (FSI) method is proposed in this paper, which can simultaneously achieve better focusing performance in both the range and cross-range domains. Experiments using simulated data and real data demonstrate the superiority of our proposed framework over existing sparsity-driven methods and range-Doppler methods.

Attentive Learning Facilitates Generalization of Neural Networks.

This article studies the generalization of neural networks (NNs) by examining how a network changes when trained on a training sample with or without out-of-distribution (OoD) examples. If the network's predictions are less influenced by fitting OoD examples, then the network learns attentively from the clean training set. A new notion, dataset-distraction stability, is proposed to measure the influence. Extensive CIFAR-10/100 experiments on the different VGG, ResNet, WideResNet, ViT architectures, and optimizers show a negative correlation between the dataset-distraction stability and generalizability. With the distraction stability, we decompose the learning process on the training set S into multiple learning processes on the subsets of S drawn from simpler distributions, i.e., distributions of smaller intrinsic dimensions (IDs), and furthermore, a tighter generalization bound is derived. Through attentive learning, miraculous generalization in deep learning can be explained and novel algorithms can also be designed.

Dual-View Learning Based on Images and Sequences for Molecular Property Prediction.

The prediction of molecular properties remains a challenging task in the field of drug design and development. Recently, there has been a growing interest in the analysis of biological images. Molecular images, as a novel representation, have proven to be competitive, yet they lack explicit information and detailed semantic richness. Conversely, semantic information in SMILES sequences is explicit but lacks spatial structural details. Therefore, in this study, we focus on and explore the relationship between these two types of representations, proposing a novel multimodal architecture named ISMol. ISMol relies on a cross-attention mechanism to extract information representations of molecules from both images and SMILES strings, thereby predicting molecular properties. Evaluation results on 14 small molecule ADMET datasets indicate that ISMol outperforms machine learning (ML) and deep learning (DL) models based on single-modal representations. In addition, we analyze our method through a large number of experiments to test the superiority, interpretability and generalizability of the method. In summary, ISMol offers a powerful deep learning toolbox for drug discovery in a variety of molecular properties.