Drug repositioning (DR) is a promising strategy to discover new indicators of approved drugs with artificial intelligence techniques, thus improving traditional drug discovery and development. However, most of DR computational methods fall short of taking into account the non-Euclidean nature of biomedical network data. To overcome this problem, a deep learning framework, namely DDAGDL, is proposed to predict drug-drug associations (DDAs) by using geometric deep learning (GDL) over heterogeneous information network (HIN). Incorporating complex biological information into the topological structure of HIN, DDAGDL effectively learns the smoothed representations of drugs and diseases with an attention mechanism. Experiment results demonstrate the superior performance of DDAGDL on three real-world datasets under 10-fold cross-validation when compared with state-of-the-art DR methods in terms of several evaluation metrics. Our case studies and molecular docking experiments indicate that DDAGDL is a promising DR tool that gains new insights into exploiting the geometric prior knowledge for improved efficacy.

Identifying new indications for drugs plays an essential role at many phases of drug research and development. Computational methods are regarded as an effective way to associate drugs with new indications. However, most of them complete their tasks by constructing a variety of heterogeneous networks without considering the biological knowledge of drugs and diseases, which are believed to be useful for improving the accuracy of drug repositioning. To this end, a novel heterogeneous information network (HIN) based model, namely HINGRL, is proposed to precisely identify new indications for drugs based on graph representation learning techniques. More specifically, HINGRL first constructs a HIN by integrating drug–disease, drug–protein and protein–disease biological networks with the biological knowledge of drugs and diseases. Then, different representation strategies are applied to learn the features of nodes in the HIN from the topological and biological perspectives. Finally, HINGRL adopts a Random Forest classifier to predict unknown drug–disease associations based on the integrated features of drugs and diseases obtained in the previous step. Experimental results demonstrate that HINGRL achieves the best performance on two real datasets when compared with state-of-the-art models. Besides, our case studies indicate that the simultaneous consideration of network topology and biological knowledge of drugs and diseases allows HINGRL to precisely predict drug–disease associations from a more comprehensive perspective. The promising performance of HINGRL also reveals that the utilization of rich heterogeneous information provides an alternative view for HINGRL to identify novel drug–disease associations especially for new diseases.

Performing an accurate clustering analysis is of great significance for us to understand the behavior of complex networks, and a variety of graph clustering algorithms have thus been proposed to do so by taking into account network topology and node attributes. Among them, FCAN is an established fuzzy clustering algorithm that optimizes the memberships of nodes based on dense structures and content relevance. This paper proposes an improved fuzzy-based graph clustering algorithm, namely FCAN-MOPSO, that retains all the benefits associated with FCAN while achieving significantly increased convergence rate using multi-objective particle swarm optimization (MOPSO). To do so, FCAN-MOPSO first modifies the original optimization model of FCAN by adopting an instance-frequency-weighted regularization, which enhances the ability of FCAN-MOPSO to handle the imbalance observed in the distribution of fuzzy memberships of nodes. After that, FCAN-MOPSO decomposes its optimization problem into a set of sub-optimization problems. Following the MOPSO framework, FCAN-MOPSO develops an effective solution to reach a consensus optimization among them by balancing the global exploration and local exploitation abilities of particles. A theoretical analysis is provided to prove the global convergence of FCAN-MOPSO. Extensive experiments have been conducted to evaluate the performance of FCAN-MOPSO on five real-world complex networks with different scale, and experimental results demonstrate that when compared with state-of-the-art clustering algorithms, FCAN-MOPSO achieves a better accuracy performance with improved convergence. Hence, FCAN-MOPSO is a promising graph clustering algorithm to precisely and efficiently discover clusters in complex networks.

Drug–drug interactions (DDIs) are known as the main cause of life-threatening adverse events, and their identification is a key task in drug development. Existing computational algorithms mainly solve this problem by using advanced representation learning techniques. Though effective, few of them are capable of performing their tasks on biomedical knowledge graphs (KGs) that provide more detailed information about drug attributes and drug-related triple facts. In this work, an attention-based KG representation learning framework, namely DDKG, is proposed to fully utilize the information of KGs for improved performance of DDI prediction. In particular, DDKG first initializes the representations of drugs with their embeddings derived from drug attributes with an encoder–decoder layer, and then learns the representations of drugs by recursively propagating and aggregating first-order neighboring information along top-ranked network paths determined by neighboring node embeddings and triple facts. Last, DDKG estimates the probability of being interacting for pairwise drugs with their representations in an end-to-end manner. To evaluate the effectiveness of DDKG, extensive experiments have been conducted on two practical datasets with different sizes, and the results demonstrate that DDKG is superior to state-of-the-art algorithms on the DDI prediction task in terms of different evaluation metrics across all datasets.

Motivation: Clustering analysis in a biological network is to group biological entities into functional modules, thus providing valuable insight into the understanding of complex biological systems. Existing clustering techniques make use of lower-order connectivity patterns at the level of individual biological entities and their connections, but few of them can take into account of higher-order connectivity patterns at the level of small network motifs. Results: Here, we present a novel clustering framework, namely HiSCF, to identify functional modules based on the higher-order structure information available in a biological network. Taking advantage of higher-order Markov stochastic process, HiSCF is able to perform the clustering analysis by exploiting a variety of network motifs. When compared with several state-of-the-art clustering models, HiSCF yields the best performance for two practical clustering applications, i.e. protein complex identification and gene co-expression module detection, in terms of accuracy. The promising performance of HiSCF demonstrates that the consideration of higher-order network motifs gains new insight into the analysis of biological networks, such as the identification of overlapping protein complexes and the inference of new signaling pathways, and also reveals the rich higher-order organizational structures presented in biological networks. Availability and implementation: HiSCF is available at https://github.com/allenv5/HiSCF. Supplementary information: Supplementary data are available at Bioinformatics online.

High-dimensional and sparse (HiDS) matrices are commonly seen in big-data-related industrial applications like recommender systems. Latent factor (LF) models have proven to be accurate and efficient in extracting hidden knowledge from them. However, they mostly fail to fulfill the non-negativity constraints that describe the non-negative nature of many industrial data. Moreover, existing models suffer from slow convergence rate. An alternating-direction-method of multipliers-based non-negative LF (AMNLF) model decomposes the task of non-negative LF analysis on an HiDS matrix into small subtasks, where each task is solved based on the latest solutions to the previously solved ones, thereby achieving fast convergence and high prediction accuracy for its missing data. This paper theoretically analyzes the characteristics of an AMNLF model, and presents detailed empirical studies regarding its performance on nine HiDS matrices from industrial applications currently in use. Therefore, its capability of addressing HiDS matrices is justified in both theory and practice.