RESUMEN
Molecular property prediction (MPP) is a crucial and fundamental task for AI-aided drug discovery (AIDD). Recent studies have shown great promise of applying self-supervised learning (SSL) to producing molecular representations to cope with the widely-concerned data scarcity problem in AIDD. As some specific substructures of molecules play important roles in determining molecular properties, molecular representations learned by deep learning models are expected to attach more importance to such substructures implicitly or explicitly to achieve better predictive performance. However, few SSL pre-trained models for MPP in the literature have ever focused on such substructures. To challenge this situation, this paper presents a Chemistry-Aware Fragmentation for Effective MPP (CAFE-MPP in short) under the self-supervised contrastive learning framework. First, a novel fragment-based molecular graph (FMG) is designed to represent the topological relationship between chemistry-aware substructures that constitute a molecule. Then, with well-designed hard negative pairs, a is pre-trained on fragment-level by contrastive learning to extract representations for the nodes in FMGs. Finally, a Graphormer model is leveraged to produce molecular representations for MPP based on the embeddings of fragments. Experiments on 11 benchmark datasets show that the proposed CAFE-MPP method achieves state-of-the-art performance on 7 of the 11 datasets and the second-best performance on 3 datasets, compared with six remarkable self-supervised methods. Further investigations also demonstrate that CAFE-MPP can learn to embed molecules into representations implicitly containing the information of fragments highly correlated to molecular properties, and can alleviate the over-smoothing problem of graph neural networks.
Asunto(s)
Benchmarking , Descubrimiento de Drogas , Redes Neurales de la Computación , Aprendizaje Automático SupervisadoRESUMEN
MOTIVATION: Contrastive learning has been widely used as pretext tasks for self-supervised pre-trained molecular representation learning models in AI-aided drug design and discovery. However, existing methods that generate molecular views by noise-adding operations for contrastive learning may face the semantic inconsistency problem, which leads to false positive pairs and consequently poor prediction performance. RESULTS: To address this problem, in this article, we first propose a semantic-invariant view generation method by properly breaking molecular graphs into fragment pairs. Then, we develop a Fragment-based Semantic-Invariant Contrastive Learning (FraSICL) model based on this view generation method for molecular property prediction. The FraSICL model consists of two branches to generate representations of views for contrastive learning, meanwhile a multi-view fusion and an auxiliary similarity loss are introduced to make better use of the information contained in different fragment-pair views. Extensive experiments on various benchmark datasets show that with the least number of pre-training samples, FraSICL can achieve state-of-the-art performance, compared with major existing counterpart models. AVAILABILITY AND IMPLEMENTATION: The code is publicly available at https://github.com/ZiqiaoZhang/FraSICL.
Asunto(s)
Benchmarking , Semántica , Modelos MolecularesRESUMEN
Self-supervised pretrained models are gaining increasingly more popularity in AI-aided drug discovery, leading to more and more pretrained models with the promise that they can extract better feature representations for molecules. Yet, the quality of learned representations has not been fully explored. In this work, inspired by the two phenomena of Activity Cliffs (ACs) and Scaffold Hopping (SH) in traditional Quantitative Structure-Activity Relationship analysis, we propose a method named Representation-Property Relationship Analysis (RePRA) to evaluate the quality of the representations extracted by the pretrained model and visualize the relationship between the representations and properties. The concepts of ACs and SH are generalized from the structure-activity context to the representation-property context, and the underlying principles of RePRA are analyzed theoretically. Two scores are designed to measure the generalized ACs and SH detected by RePRA, and therefore, the quality of representations can be evaluated. In experiments, representations of molecules from 10 target tasks generated by 7 pretrained models are analyzed. The results indicate that the state-of-the-art pretrained models can overcome some shortcomings of canonical Extended-Connectivity FingerPrints, while the correlation between the basis of the representation space and specific molecular substructures are not explicit. Thus, some representations could be even worse than the canonical fingerprints. Our method enables researchers to evaluate the quality of molecular representations generated by their proposed self-supervised pretrained models. And our findings can guide the community to develop better pretraining techniques to regularize the occurrence of ACs and SH.
Asunto(s)
Fármacos Anti-VIH , Descubrimiento de Drogas , Hidrolasas , Aprendizaje , Relación Estructura-Actividad CuantitativaRESUMEN
MOTIVATION: Accurately predicting drug-target interaction (DTI) is a crucial step to drug discovery. Recently, deep learning techniques have been widely used for DTI prediction and achieved significant performance improvement. One challenge in building deep learning models for DTI prediction is how to appropriately represent drugs and targets. Target distance map and molecular graph are low dimensional and informative representations, which however have not been jointly used in DTI prediction. Another challenge is how to effectively model the mutual impact between drugs and targets. Though attention mechanism has been used to capture the one-way impact of targets on drugs or vice versa, the mutual impact between drugs and targets has not yet been explored, which is very important in predicting their interactions. RESULTS: Therefore, in this article we propose MINN-DTI, a new model for DTI prediction. MINN-DTI combines an interacting-transformer module (called Interformer) with an improved Communicative Message Passing Neural Network (CMPNN) (called Inter-CMPNN) to better capture the two-way impact between drugs and targets, which are represented by molecular graph and distance map, respectively. The proposed method obtains better performance than the state-of-the-art methods on three benchmark datasets: DUD-E, human and BindingDB. MINN-DTI also provides good interpretability by assigning larger weights to the amino acids and atoms that contribute more to the interactions between drugs and targets. AVAILABILITY AND IMPLEMENTATION: The data and code of this study are available at https://github.com/admislf/MINN-DTI.
Asunto(s)
Redes Neurales de la Computación , Proteínas , Humanos , Proteínas/química , Simulación por Computador , Desarrollo de Medicamentos/métodos , Descubrimiento de Drogas/métodosRESUMEN
MOTIVATION: Molecular property prediction is a hot topic in recent years. Existing graph-based models ignore the hierarchical structures of molecules. According to the knowledge of chemistry and pharmacy, the functional groups of molecules are closely related to its physio-chemical properties and binding affinities. So, it should be helpful to represent molecular graphs by fragments that contain functional groups for molecular property prediction. RESULTS: In this article, to boost the performance of molecule property prediction, we first propose a definition of molecule graph fragments that may be or contain functional groups, which are relevant to molecular properties, then develop a fragment-oriented multi-scale graph attention network for molecular property prediction, which is called FraGAT. Experiments on several widely used benchmarks are conducted to evaluate FraGAT. Experimental results show that FraGAT achieves state-of-the-art predictive performance in most cases. Furthermore, our case studies show that when the fragments used to represent the molecule graphs contain functional groups, the model can make better predictions. This conforms to our expectation and demonstrates the interpretability of the proposed model. AVAILABILITY AND IMPLEMENTATION: The code and data underlying this work are available in GitHub, at https://github.com/ZiqiaoZhang/FraGAT. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
RESUMEN
At the early stages of the drug discovery, molecule toxicity prediction is crucial to excluding drug candidates that are likely to fail in clinical trials. In this paper, we presented a novel molecular representation method and developed a corresponding deep learning-based framework called TOP (the abbreviation of TOxicity Prediction). TOP integrates specifically designed data preprocessing methods, an RNN based on bidirectional gated recurrent unit (BiGRU), and fully connected neural networks for end-to-end molecular representation learning and chemical toxicity prediction. TOP can automatically learn a mixed molecular representation from not only SMILES contextual information that describes the molecule structure, but also physiochemical properties. Therefore, TOP can overcome the drawbacks of existing methods that use either of them, thus greatly promotes toxicity prediction accuracy. We conducted extensive experiments over 14 classic toxicity prediction tasks on three different benchmark datasets, including balanced and imbalanced ones. The results show that, with the help of the novel molecular representation method, TOP significantly outperforms not only three baseline machine learning methods, but also five state-of-the-art methods.
Asunto(s)
Quimioinformática/métodos , Aprendizaje Profundo , Descubrimiento de Drogas/métodos , Farmacología Clínica/métodos , Pruebas de Toxicidad/métodos , Conjuntos de Datos como Asunto , Descubrimiento de Drogas/estadística & datos numéricos , Predicción/métodos , Humanos , Farmacología Clínica/estadística & datos numéricos , Pruebas de Toxicidad/estadística & datos numéricosRESUMEN
Glaucoma is an eye disease with a high rate of blindness and a complex pathogenesis. Ocular hypertension (OHT) is a critical risk factor, and retinal ischemia/reperfusion (I/R) is an important pathophysiological basis. This study was designed to investigate the retinal neuroprotective effect of oral naringenin in an acute retinal I/R model and a chronic OHT model and the possible mechanism involved. After the I/R and OHT models were established, mice were given vehicle or naringenin (100 mg/kg or 300 mg/kg). Hematoxylin-eosin (HE) staining and immunostaining of RBPMS and glial fibrillary acidic protein (GFAP) were used to evaluate retinal injury. GFAP, CD38, Sirtuin1 (SIRT1), and NOD-like receptor protein 3 (NLRP3) expression levels were measured by Western blotting. In the OHT model, intraocular pressure (IOP) was dynamically maintained at approximately 20-25 mmHg after injury. The retinal structure was damaged, and retinal ganglion cells (RGCs) were lost in both models. Naringenin ameliorated the abovementioned indications but also demonstrated that high concentrations of naringenin significantly inhibited retinal astrocyte activation and inhibited damage-induced increases in the expression of GFAP, NLRP3, and CD38 proteins, while SIRT1 protein expression was upregulated. This study showed for the first time that naringenin can reduce microbead-induced IOP elevation in the OHT model, providing new evidence for the application of naringenin in glaucoma. Naringenin may mediate the CD38/SIRT1 signaling pathway, inhibit astrocyte activation, and ultimately exert an anti-inflammatory effect to achieve retinal neuroprotection.
Asunto(s)
Glaucoma , Hipertensión Ocular , Enfermedades de la Retina , Ratones , Animales , Flavonoides , Sirtuina 1 , Proteína con Dominio Pirina 3 de la Familia NLR , Glaucoma/metabolismo , Hipertensión Ocular/patología , Enfermedades de la Retina/metabolismo , Presión Intraocular , Modelos Animales de EnfermedadRESUMEN
While significant advances have been made in predicting static protein structures, the inherent dynamics of proteins, modulated by ligands, are crucial for understanding protein function and facilitating drug discovery. Traditional docking methods, frequently used in studying protein-ligand interactions, typically treat proteins as rigid. While molecular dynamics simulations can propose appropriate protein conformations, they're computationally demanding due to rare transitions between biologically relevant equilibrium states. In this study, we present DynamicBind, a deep learning method that employs equivariant geometric diffusion networks to construct a smooth energy landscape, promoting efficient transitions between different equilibrium states. DynamicBind accurately recovers ligand-specific conformations from unbound protein structures without the need for holo-structures or extensive sampling. Remarkably, it demonstrates state-of-the-art performance in docking and virtual screening benchmarks. Our experiments reveal that DynamicBind can accommodate a wide range of large protein conformational changes and identify cryptic pockets in unseen protein targets. As a result, DynamicBind shows potential in accelerating the development of small molecules for previously undruggable targets and expanding the horizons of computational drug discovery.
Asunto(s)
Simulación de Dinámica Molecular , Proteínas , Ligandos , Proteínas/metabolismo , Conformación Proteica , Descubrimiento de Drogas , Unión Proteica , Simulación del Acoplamiento MolecularRESUMEN
The current evaluation of M&A performance lacks consideration of M&A motives. In this paper, we theoretically analyse and empirically test the effect of network synergy generated by M&A on the degree of realization of corporate M&A motives and the mechanism of its effect by constructing an equity network between a listed company and its subsidiaries within the company. The results show that the greater the variation of internal network node degree and strength, the more beneficial it is to promote the degree of realization of corporate M&A motivation; the results of further mechanism of action tests show that the variation of network node degree and strength has significant effects on economies of scale, economies of scope, and transaction costs; Furthermore, the heterogeneity test finds that the effect of variation of network node degree and strength to promote the degree of realization of corporate M&A motivation is more significant in the case of non-cash payment method and related M&A. This paper extends the study of complex networks to the field of M&A and uniquely explains the paradox of the "high failure rate" of M&A and the increasing activity of M&A activities from the perspective of network synergy, which helps to rationalize the M&A behavior of enterprises and further regulate the M&A behavior of listed companies by regulatory authorities.
Asunto(s)
Motivación , ChinaRESUMEN
In this paper, a distributed cooperative filtering strategy for state estimation has been developed for mobile sensor networks in a spatial-temporal varying field modeled by the advection-diffusion equation. Sensors are organized into distributed cells that resemble a mesh grid covering a spatial area, and estimation of the field value and gradient information at each cell center is obtained by running a constrained cooperative Kalman filter while incorporating the sensor measurements and information from neighboring cells. Within each cell, the finite volume method is applied to discretize and approximate the advection-diffusion equation. These approximations build the weakly coupled relationships between neighboring cells and define the constraints that the cooperative Kalman filters are subjected to. With the estimated information, a gradient-based formation control law has been developed that enables the sensor network to adjust formation size by utilizing the estimated gradient information. Convergence analysis has been conducted for both the distributed constrained cooperative Kalman filter and the formation control. Simulation results with a 9-cell 12-sensor network validate the proposed distributed filtering method and control law.
RESUMEN
Glaucoma is a kind of optic neuropathy mainly manifested in the permanent death of retinal ganglion cells (RGCs), atrophy of the optic nerve, and loss of visual ability. The main risk factors for glaucoma consist of the pathological elevation of intraocular pressure (IOP) and aging. Although the mechanism of glaucoma remains an open question, a theory related to mitochondrial dysfunction has been emerging in the last decade. Reactive oxygen species (ROS) from the mitochondrial respiratory chain are abnormally produced as a result of mitochondrial dysfunction. Oxidative stress takes place when the cellular antioxidant system fails to remove excessive ROS promptly. Meanwhile, more and more studies show that there are other common features of mitochondrial dysfunction in glaucoma, including damage of mitochondrial DNA (mtDNA), defective mitochondrial quality control, ATP reduction, and other cellular changes, which are worth summarizing and further exploring. The purpose of this review is to explore mitochondrial dysfunction in the mechanism of glaucomatous optic neuropathy. Based on the mechanism, the existing therapeutic options are summarized, including medications, gene therapy, and red-light therapy, which are promising to provide feasible neuroprotective ideas for the treatment of glaucoma.
RESUMEN
Transcriptional programming leverages systems of engineered transcription factors to impart decision-making (e.g., Boolean logic) in chassis cells. The number of components used to construct said decision-making systems is rapidly increasing, making an exhaustive experimental evaluation of iterations of biological circuits impractical. Accordingly, we posited that a predictive tool is needed to guide and accelerate the design of transcriptional programs. The work described here involves the development and experimental characterization of a large collection of network-capable single-INPUT logical operationsâi.e., engineered BUFFER (repressor) and engineered NOT (antirepressor) logical operations. Using this single-INPUT data and developed metrology, we were able to model and predict the performances of all fundamental two-INPUT compressed logical operations (i.e., compressed AND gates and compressed NOR gates). In addition, we were able to model and predict the performance of compressed mixed phenotype logical operations (A NIMPLY B gates and complementary B NIMPLY A gates). These results demonstrate that single-INPUT data is sufficient to accurately predict both the qualitative and quantitative performance of a complex circuit. Accordingly, this work has set the stage for the predictive design of transcriptional programs of greater complexity.
Asunto(s)
Lógica , Factores de Transcripción , Factores de Transcripción/genéticaRESUMEN
Purpose: The purpose of this study is to investigate the anti-pyroptotic effect of resveratrol in the context of ischemia-reperfusion (I/R)-induced retinal injury, with a particular focus on Müller glial cells (MGCs) and to elucidate the underlying molecular mechanisms. Methods: The retinal I/R model was constructed in mice and pyroptotic markers were measured at six, 12, 24, 48, and 72 hours after I/R injury to determine the peak of pyroptotic activity. The effects of resveratrol on pyroptosis, inflammasomes, and the activation of MGCs after I/R injury were observed on the retina of mice. Moreover, induction of pyroptosis in rat Müller glial cells (r-MC) via lipopolysaccharide was used to explore the effects of resveratrol on pyroptosis of r-MC in vitro. Results: After the induction of retinal I/R injury in mice, the intricate involvement of pyroptosis in the progressive degeneration of the retina was observed, reaching its zenith at the onset of 24 hours after I/R injury. Resveratrol treatment alleviated I/R injury on the retina, relieved retinal ganglion cells death. In addition, resveratrol inhibited Caspase-1 activation, gasdermin D (GSDMD-N) cleavage, the inflammasome assembly, and the release of inflammatory cytokines, simultaneously relieving the MGCs activation. Furthermore, resveratrol inhibited the pyroptosis-related NLRP3/GSDMD-N/TMS1/ASC/Caspase-1/IL-1ß pathway in r-MC cells, and mitigated cells death in vitro. Conclusions: Pyroptosis plays an important role in the pathogenesis of retinal I/R injury. Resveratrol can attenuate pyroptotic-driven damage in the retina and MGC by inhibiting the NLRP3/GSDMD-N/TMS1/ASC/Caspase-1/IL-1ß pyroptosis pathway.
Asunto(s)
Daño por Reperfusión , Resveratrol , Retina , Animales , Ratones , Ratas , Caspasa 1/metabolismo , Gasderminas , Inflamasomas/metabolismo , Interleucina-1beta/metabolismo , Proteína con Dominio Pirina 3 de la Familia NLR/metabolismo , Piroptosis , Daño por Reperfusión/tratamiento farmacológico , Daño por Reperfusión/prevención & control , Daño por Reperfusión/metabolismo , Resveratrol/farmacología , Retina/metabolismoRESUMEN
Online Social Networks generate a prodigious wealth of real-time information at an incessant rate. In this paper we study the empirical data that crawled from Twitter to describe the topology and information spreading dynamics of Online Social Networks. We propose a measurement with three measures to state the efforts of users on Twitter to get their information spreading, based on the unique mechanisms for information retransmission on Twitter. It is noticed that small fraction of users with special performance on participation can gain great influence, while most other users play a role as middleware during the information propagation. Thus a community analysis is performed and four categories of users are found with different kinds of participation that cause the information dissemination dynamics. These suggest that exiting topological measures alone may reflect little about the influence of individuals and provide new insights for information spreading.