In recent years, the "drug-target-disease" association has become increasingly complex and data-scarce. Existing methods are limited by information loss caused by explicit graph construction when modeling triplet relationships, making it difficult to effectively capture geometric structures and long-range dependencies, thereby affecting prediction performance. To address this, this paper proposes a new method based on point cloud modeling-TriCloud. This method represents each triad as a spatial point cloud, encoding semantic and topological relationships through geometric coordinates, and performs feature learning directly on an unordered point set, avoiding reliance on predefined graph structures. Based on the PointNet architecture, it introduces a multi-view feature extraction and fusion mechanism to enhance the modeling capability of global structures and complex spatial patterns. Experimental results show that TriCloud significantly outperforms existing methods on multiple benchmark datasets, achieving an AUC of 0.9995 and an AUPR of 0.9996, with all metrics ranking first. External validation demonstrates its excellent generalization ability. Feature analysis reveals that the geometric-semantic joint features of positive samples play a dominant role in classification. This study provides an efficient and reliable computational framework for drug repurposing, contributing to the development of precision medicine.

Drug repurposing relies critically on the accurate prediction of drug-target interactions (DTIs). Conventional graph neural network approaches typically model drugs and proteins as isolated nodes, focusing solely on intrinsic attributes such as molecular structure or sequence information. As a result, they often fail to capture complex synergistic effects-such as multi-target regulation. To overcome this limitation, this paper proposes the Topology-Aggregation Enhanced Graph Interaction Network (TAGIN-DTI), a novel framework that shifts the prediction paradigm from a conventional node-level view to a subnetwork-level perspective. The model aggregates topological association features from drug-drug and protein-protein interaction networks via a dual-path Transformer encoder, integrates multi-scale global and local information through a gated fusion mechanism, and incorporates MinHash-based subgraph structure encoding to enhance neighborhood topological representation. Experimental results demonstrate that TAGIN-DTI outperforms existing methods in both prediction accuracy and generalization capability, offering valuable insights for drug repurposing and target discovery.

Drug repositioning is a strategy to identify new therapeutic uses for existing drugs, significantly reducing development costs and time. Although deep learning methods for drug-target interaction prediction have advanced, most models are limited to binary relationships and struggle to capture complex ternary associations among drugs, targets, and diseases. Additionally, limitations in graph structure modeling and node feature representation often constrain their generalization capability. To address these challenges, this paper proposes GraphTransHGN, a framework integrating graph embedding, Transformer feature extraction, and heterogeneous graph neural networks. First, Node2Vec constructs graph representations of drugs and targets, with the Transformer extracting high-level semantic features. Drug-target pairs are then combined into composite nodes (DT_node), and target-disease data are incorporated to form a heterogeneous graph. Finally, the HGTConv models multi-type relationships between DT_node and diseases, enabling end-to-end prediction of potential therapeutic associations. Experimental results demonstrate that GraphTransHGN outperforms mainstream methods across key metrics, achieving an AUC of 0.9916, Recall of 0.9959, and AUPR of 0.9849, which confirms its discriminative power and robustness. The model not only improves prediction accuracy for drug repositioning but also offers a novel technical and theoretical foundation for mechanism-based drug discovery.

BACKGROUND: Drug-drug interactions (DDIs) frequently occur in combination therapy and may cause adverse effects or reduced efficacy. Existing computational approaches often fail to capture both the semantic information in drug sequences and the structural properties of drug molecules, limiting predictive power. RESULTS: We propose MDG-DDI, a deep learning framework that integrates a Frequent Consecutive Subsequence (FCS)-based Transformer encoder with a Deep Graph Network (DGN) to extract complementary semantic and structural features. These representations are fused and fed into a Graph Convolutional Network (GCN) for DDI prediction. Experiments on three benchmark datasets under transductive and inductive settings show that MDG-DDI consistently outperforms state-of-the-art methods, with particularly strong gains when predicting interactions involving unseen drugs. CONCLUSION: By jointly modeling substructure-level semantics and molecular graph structure, MDG-DDI achieves robust and accurate DDI prediction. The framework demonstrates improved generalization and offers potential for enhancing drug safety assessment and discovery.

Large language models (LLMs) such as ChatGPT have recently demonstrated significant potential in mathematical abilities, providing a valuable reasoning paradigm consistent with human natural language. However, LLMs currently have difficulty in bridging perception, language understanding, and reasoning (PLR) capabilities due to incompatibility of the underlying information flow among them, making their reasoning ability not fully elicited and challenging to accomplish complicated reasoning tasks autonomously. To resolve the above problem, a novel method called ChatABL is proposed by integrating LLMs into an abductive learning (ABL) framework, capable of unifying the three abilities effectively in a more user-friendly and understandable manner. Initially, the proposed method uses LLMs to correct the incomplete logical facts for optimizing the perception module, by summarizing and reorganizing domain knowledge represented in natural language format. Then, the perception module also provides necessary logical reasoning materials for feeding LLMs. Finally, these parts are integrated into a dynamic closed-loop system by introducing the feedback form and automatic learning strategies to mutually promote their performance. As a testbed, the variable-length handwritten equation decipherment (HED), an abstract expression of the Mayan calendar decoding, is used to demonstrate that ChatABL has reasoning ability beyond most existing state-of-the-art methods, which has been well-supported by comparative studies. To the best of authors' knowledge, the proposed ChatABL is the first attempt to explore a possible and novel avenue to approaching human-level cognitive ability via natural language interaction by means of ChatGPT.

Drug repositioning is a vital area of biomedicine, where confirming interactions between drugs and specific targets is essential for establishing the efficacy of pharmaceutical agents. Traditional in vitro screening methods have limitations, prompting the use of computer simulations as an effective alternative for predicting drug-target interactions (DTI). This approach has gained significant attention in the scientific community. In this study, we introduce MFCM-DTI, a DTI prediction model that employs multimodal features to accurately capture the intricate interactions between drug molecular structures and key amino acids of target proteins. Our results demonstrate that MFCM-DTI outperforms existing models in prediction accuracy and robustness. Furthermore, MFCM-DTI has been successfully used to predict interactions between key SARS-CoV-2 proteins and existing drugs, providing a solid foundation for developing therapeutic agents against SARS-CoV-2 infection. This study underscores the broad applicability and strong predictive capabilities of MFCM-DTI in drug-protein interaction prediction, opening new avenues for research. The predicted drug targets and interaction data offer valuable insights for future experimental validation and clinical trials, potentially driving innovation in the biomedical field.

Bioinformatics is a rapidly growing field that involves the application of computational methods to analyze and interpret biological data. One important task in bioinformatics is predicting the drug-target affinity (DTA), which plays a significant role in drug discovery through virtual screening. Effectively predicting the association between drug molecules and target molecules can speed up the drug discovery process. The DTA can be quantitatively measured. This quantifiable affinity is more precise than a simple binary relationship. In this study, we propose a deep learning model for DTA prediction that utilizes the encoder module of the Transformer architecture. Our proposed model utilizes Convolutional Neural Networks (CNNs) and the encoder module of Transformer to characterize protein and drug sequences. The model outperforms some methods such as KronRLS, SimBoost and DeepDTA as evidenced by superior evaluation metrics such as Mean Squared Error (MSE), Concordance Index (CI), and Regression toward the Mean Index $\left( {r_m^2} \right)$. These results demonstrate the effectiveness of the Transformer’s encoder in extracting meaningful representations from sequences, thereby improving the accuracy of DTA prediction. Deep learning models for DTA prediction can accelerate drug discovery by identifying drug candidates with high binding affinity to specific targets. Compared to traditional methods, the use of machine learning technology enables more effective and efficient drug discovery.

The decomposition of multi-subject fMRI data using rank-(L,L,1,1) block term decomposition (BTD) can preserve higher-way data structure and is more robust to noise effects by decomposing shared spatial maps (SMs) into a product of two rank-L loading matrices. However, since the number of whole-brain voxels is very large and rank L is larger than 1, the rank-(L,L,1,1) BTD requires high computation and memory. Therefore, we propose an accelerated rank-(L,L,1,1) BTD algorithm based upon the method of alternating least squares (ALS). We speed up updates of loading matrices by reducing fMRI data into subspaces, and add an orthonormality constraint on shared SMs to improve the performance. Moreover, we evaluate the rank-L effect on the proposed method for actual task-related fMRI data. The proposed method shows better performance when L=35. Meanwhile, experimental comparison results verify that the proposed method largely reduced (17.36 times) computation time compared to ALS while also providing satisfying separation performance.

In 1990, Hendry conjectured that all Hamiltonian chordal graphs are cycle extendable. After a series of papers confirming the conjecture for a number of graph classes, the conjecture is yet refuted by Lafond and Seamone in 2015. Given that their counterexamples are not strongly chordal graphs and they are all only 2-connected, Lafond and Seamone asked the following two questions: (1) Are Hamiltonian strongly chordal graphs cycle extendable? (2) Is there an integer $k$ such that all $k$-connected Hamiltonian chordal graphs are cycle extendable? Later, a conjecture stronger than Hendry's is proposed. In this paper, we resolve all these questions in the negative. On the positive side, we add to the list of cycle-extendable graphs two more graph classes, namely, Hamiltonian 4-fan-free chordal graphs, where every induced $K_5 - e$ has true twins, and Hamiltonian $\{4{\sc -fan}, \overline{A} \}$-free chordal graphs.

As a successful application of edge computing in the industrial production environment, prolonging the smart devices' (SDs') battery lifetime has become an important issue. In some special practical applications, the uploaded data from SDs to vehicle base stations (VBSs) or servers can be merged between SDs with a fixed size, which is called superposed data. In this article, we consider the superposed data uploading problem in a decentralized device-to-device communication system. The task of the problem is to minimize the total energy consumption of uploading data. We reduce it into a combinatorial optimization problem from the graph theory perspective. For VBSs or servers with infinite capacities, we propose an optimal algorithm with polynomial running time. When VBSs or servers have limited capacities, the problem is NP-hard even in very special cases. For this NP-hard problem, we give two heuristic algorithms and the corresponding numerical simulation results.

We study the Maximum Satisfiability problem (MaxSAT). Particularly, we derive a branching algorithm of running time O*(1.2989^m) for the MaxSAT problem, where m denotes the number of clauses in the given CNF formula. Our algorithm considerably improves the previous best result O*(1.3248^m) by Chen and Kanj [2004] published 15 years ago. For our purpose, we derive improved branching strategies for variables of degrees 3, 4, and 5. The worst case of our branching algorithm is at variables of degree 4 which occur twice both positively and negatively in the given CNF formula. To serve the branching rules and shrink the size of the CNF formula, we also propose a variety of reduction rules which can be exhaustively applied in polynomial time and, moreover, some of them solve a bottleneck of the previous best algorithm.

Location estimation technology under the wireless environment has become a vital technology in the field of mobile edge computing. Especially, under the mobile edge of entire networks environment, indoor location estimation is gradually getting the interest research and application topic, due to technical constraints of global positioning system technology for indoor environment and the popularity of the mobile edge computing servers. In this paper, the widely used single-model framework for indoor localization is presented as an introduction, which consists of three stages: 1) sample data collection; 2) model building; and 3) localization estimation. And then, through analyzing of the actual scene of indoor localization, a new framework for indoor localization under mobile edge computing environment, named Multimodel, is proposed from the theoretical perspective. It is mainly based on the observation that the environment of the sample data collection and that of localization data collection may change seriously. In order to make up for the shortcomings of this framework, two combinatorial optimization problems are proposed. Later, we discuss the NP-hardness of them in several different cases. In addition, two heuristic algorithms are given, and the performance of which are illustrated by the corresponding experimental results.