Drug-induced liver injury (DILI) has been the single most frequent cause of safety-related drug marketing withdrawals for past 50 years. Recently, deep learning (DL) successfully applied in many fields due to its exceptional and automatic ability. In this study, DILI prediction models were developed using DL architectures, best model trained on 475 drugs predicted an external validation set 198 with accuracy 86.9%, sensitivity 82.5%, specificity 92.9%, area under curve 0.955, which is better...

CavityPlus is a web server that offers protein cavity detection and various functional analyses. Using three-dimensional structural information as the input, applies CAVITY to detect potential binding sites on surface of given structure rank them based ligandability druggability scores. These can be further analysed using three submodules, CavPharmer, CorrSite, CovCys. CavPharmer uses receptor-based pharmacophore modelling program, Pocket, automatically extract features within cavities....

Median lethal death, LD50, is a general indicator of compound acute oral toxicity (AOT). Various in silico methods were developed for AOT prediction to reduce costs and time. In this study, we an improved molecular graph encoding convolutional neural networks (MGE-CNN) architecture construct three types high-quality models: regression model (deepAOT-R), multiclassification (deepAOT-C), multitask (deepAOT-CR). These predictive models highly outperformed previously reported models. For the two...

Retrosynthetic route planning can be considered a rule-based reasoning procedure. The possibilities for each transformation are generated based on collected reaction rules, and then potential routes recommended by various optimization algorithms. Although there has been much progress in computer-assisted retrosynthetic prediction, fully data-driven automatic remains challenging. Here we present template-free approach that is independent of templates, or atom mapping, to implement planning....

The liquid-liquid phase separation (LLPS) of biomolecules in cell underpins the formation membraneless organelles, which are condensates protein, nucleic acid, or both, and play critical roles cellular function. Dysregulation LLPS is implicated a number diseases. Although has been investigated intensively recent years, knowledge prevalence distribution proteins (PSPs) still lag behind. Development computational methods to predict PSPs therefore great importance for comprehensive...

Adverse side effects of drug-drug interactions induced by human cytochrome P450 (CYP450) inhibition is an important consideration in drug discovery. It highly desirable to develop computational models that can predict the inhibitive effect a compound against specific CYP450 isoform. In this study, we developed multitask model for concurrent prediction five major isoforms, namely, 1A2, 2C9, 2C19, 2D6, and 3A4. The was built training autoencoder deep neural network (DNN) on large dataset...

Optical chirality occurs when materials interact differently with light in a specific circular polarization state. Chiroptical phenomena inspire wide interdisciplinary investigations, which require advanced designs to reach strong for practical applications. The development of artificial intelligence provides new vision the manipulation light-matter interaction beyond theoretical interpretation. Here, we report self-consistent framework named Bayesian optimization and convolutional neural...

Abstract The COVID-19 pandemic calls for rapid development of effective treatments. Although various drug repurpose approaches have been used to screen the FDA-approved drugs and candidates in clinical phases against SARS-CoV-2, coronavirus that causes this disease, no magic bullets found until now. In study, we directed message passing neural network first build a broad-spectrum anti-beta-coronavirus compound prediction model, which gave satisfactory predictions on newly reported active...

Prostate cancer (PCa) is the most prevalent among men in United States and remains second-leading cause of mortality men. Paclitaxel (PTX) first line chemotherapy for PCa treatment, but its therapeutic efficacy greatly restricted by nonspecific distribution vivo. Prostate-specific membrane antigen (PSMA) overexpressed on surface cells, expression level increases with aggressiveness, while being present at low levels normal cells. The high PSMA cells offers an opportunity target delivery...

Precisely evaluating the protein-ligand interaction is crucial in drug screening and optimization. There are significant advances application of machine learning approaches to developing scoring functions recent years. However, traditional docking softwares existing deep-learning methods remain unsolved limitations terms pose quality binding affinity prediction accuracy. Furthermore, learning-based hard generalize unseen cases due scarcity structure-affinity data, mostly lack physical...

Precisely evaluating the protein-ligand interaction is crucial in drug screening and optimization. There are significant advances application of machine learning approaches to developing scoring functions recent years. However, traditional docking softwares existing deep-learning methods remain unsolved limitations terms pose quality binding affinity prediction accuracy. Furthermore, learning-based hard generalize unseen cases due scarcity structure-affinity data, mostly lack physical...

Abstract The liquid-liquid phase separation (LLPS) of bio-molecules in cell underpins the formation membraneless organelles, which are condensates protein, nucleic acid, or both, and play critical roles cellular functions. dysregulation LLPS might be implicated a number diseases. Although biomolecules has been investigated intensively recent years, knowledge prevalence distribution proteins (PSPs) is still lag behind. Development computational methods to predict PSPs therefore great...

Allostery is an important mechanism that biological systems use to regulate function at a distant site. Allosteric drugs have attracted much attention in recent years due their high specificity and the possibility of overcoming existing drug-resistant mutations. However, discovery allosteric remains challenging as regulation mechanisms are not clearly understood sites cannot be accurately predicted. In this study, we analyzed dominant modes determine motion correlations between orthosteric...

For quantitative structure-property relationship (QSPR) studies in chemoinformatics, it is important to get interpretable between chemical properties and features. However, the predictive power interpretability of QSPR models are usually two different objectives that difficult achieve simultaneously. A deep learning architecture using molecular graph encoding convolutional neural networks (MGE-CNN) provided a universal strategy construct with high power. Instead application-specific preset...

Cytochrome P450 17A1 (CYP17A1) is associated in the steroid hormone biosynthesis human. As cell proliferation of prostate cancer response to androgen steroid, an inhibition CYP17A1 becomes alternative approach inhibit and support treatment cancer. However, biology-driven inhibitor development poorly elucidated. The aims this study are address structural differences at atomic-level between inhibitors i.e., abiraterone TOK-001, further investigate effect point mutation on active site stability...

Server-less RFID systems are used more and widespread recently, which allow readers authenticating a specific tag without the help of on-line backend servers, it brings higher design requirements for security protocols. In this paper, mutual authentication protocol its corresponding serach ptotocol server-less proposed. The properties these protocols analyzed as well by comparing with related

Deep generative models have gained significant advancements to accelerate drug discovery by generating bioactive chemicals against desired targets. Nevertheless, most generated compounds that been validated for potent bioactivity often exhibit structural novelty levels fall short of satisfaction, thereby providing limited inspiration human medicinal chemists. The challenge faced lies in their ability produce are both and novel, rather than merely making minor modifications known actives...

Except common security and privacy requirements, well scalability supporting tag ownership transfer are important requests for RFID systems. In this paper， an efficient mutual authentication protocol is proposed, only requires O(1) work to identify authenticate a in the backend server suitable low-cost The performance of proposed analyzed as well.

<div><div><div><p>We present an attention-based Transformer model for automatic retrosynthesis route planning. Our approach starts from <a></a><a>reactants prediction of single-step organic reactions gi</a>ven products, <a>followed by Monte Carlo tree search-based retrosynthetic pathway prediction</a>. Trained on two datasets the United States patent literature, our models achieved a top-1 accuracy over 54.6% and 63.0% with more...

<title>Abstract</title> We have developed MaxQsaring, a novel universal framework integrating molecular descriptors, fingerprints, and deep-learning pretrained representations, to predict the properties of compounds. Applied case study hERG (human Ether-à-go-go-Related Gene) blockage prediction, MaxQsaring achieved state-of-the-art performance on two external challenging datasets through automatic optimal feature combinations, successfully identified top 10 important interpretable features...