MOTIVATION: In recent years, circular RNAs (circRNAs), the particular form of RNA with a closed-loop structure, have attracted widespread attention due to their physiological significance (they can directly bind proteins), leading to the development of numerous protein site identification algorithms. Unfortunately, these studies are supervised and require the vast majority of labeled samples in training to produce superior performance. But the acquisition of sample labels requires a large number of biological experiments and is difficult to obtain. RESULTS: To resolve this matter that a great deal of tags need to be trained in the circRNA-binding site prediction task, a self-supervised learning binding site identification algorithm named CircSI-SSL is proposed in this article. According to the survey, this is unprecedented in the research field. Specifically, CircSI-SSL initially combines multiple feature coding schemes and employs RNA_Transformer for cross-view sequence prediction (self-supervised task) to learn mutual information from the multi-view data, and then fine-tuning with only a few sample labels. Comprehensive experiments on six widely used circRNA datasets indicate that our CircSI-SSL algorithm achieves excellent performance in comparison to previous algorithms, even in the extreme case where the ratio of training data to test data is 1:9. In addition, the transplantation experiment of six linRNA datasets without network modification and hyperparameter adjustment shows that CircSI-SSL has good scalability. In summary, the prediction algorithm based on self-supervised learning proposed in this article is expected to replace previous supervised algorithms and has more extensive application value. AVAILABILITY AND IMPLEMENTATION: The source code and data are available at https://github.com/cc646201081/CircSI-SSL.

Abstract Clustering algorithm is one of the most widely used and influential analysis techniques. With the advent of deep learning, deep embedding clustering algorithms have rapidly evolved and yield promising results. Much of the success of these algorithms depends on the potential expression captured by the autoencoder network. Therefore, the quality of the potential expression directly determines the algorithm’s performance. In view of this, researchers have proposed many improvements. Although the performance has been slightly improved, they all have one shortcoming, that is, too much emphasis is placed on the original data reconstruction ability during the process of feature expression, which greatly limits the further expression of potential features according to specific clustering tasks. Moreover, there is a large amount of noise in the original data, so blindly emphasizing reconstruction will only backfire. Hence, we innovatively propose a deep embedding clustering algorithm based on residual autoencoder (DECRA) after in-depth research. Specifically, a novel autoencoder network with residual structure is proposed and introduced into deep embedded clustering tasks. The network introduces an adaptive weight layer in feature representation z, which can make it have good robustness, generalization for specific tasks, and adaptive learning of better feature embeddings according to category classification. In this paper, the reasons for the validity of this structure are explained theoretically, and comprehensive experiments on six benchmark datasets including various types show that the clustering performance of the DECRA is very competitive and significantly superior to the most advanced methods.

BACKGROUND: Circular RNAs (circRNAs) play a significant role in some diseases by acting as transcription templates. Therefore, analyzing the interaction mechanism between circRNA and RNA-binding proteins (RBPs) has far-reaching implications for the prevention and treatment of diseases. Existing models for circRNA-RBP identification usually adopt convolution neural network (CNN), recurrent neural network (RNN), or their variants as feature extractors. Most of them have drawbacks such as poor parallelism, insufficient stability, and inability to capture long-term dependencies. METHODS: In this paper, we propose a new method completely using the self-attention mechanism to capture deep semantic features of RNA sequences. On this basis, we construct a CircSSNN model for the cirRNA-RBP identification. The proposed model constructs a feature scheme by fusing circRNA sequence representations with statistical distributions, static local contexts, and dynamic global contexts. With a stable and efficient network architecture, the distance between any two positions in a sequence is reduced to a constant, so CircSSNN can quickly capture the long-term dependencies and extract the deep semantic features. RESULTS: Experiments on 37 circRNA datasets show that the proposed model has overall advantages in stability, parallelism, and prediction performance. Keeping the network structure and hyperparameters unchanged, we directly apply the CircSSNN to linRNA datasets. The favorable results show that CircSSNN can be transformed simply and efficiently without task-oriented tuning. CONCLUSIONS: In conclusion, CircSSNN can serve as an appealing circRNA-RBP identification tool with good identification performance, excellent scalability, and wide application scope without the need for task-oriented fine-tuning of parameters, which is expected to reduce the professional threshold required for hyperparameter tuning in bioinformatics analysis.

Abstract Reversible circuit synthesis methods based on decision diagrams achieve low quantum costs but do not account for quantum bit (qubit) limits for the application of reversible logic in quantum computing. Here, a synthesis method using sub-graphs of shared functional decision diagrams (SFDDs) is proposed for reducing the number of lines when synthesizing reversible circuits. An SFDD is partitioned into sub-graphs by exploiting the longest dominant-active paths, and the sub-graphs are mapped to reversible gate cascades. To further reduce the number of lines, template root matching is presented for reusing circuit lines. Experimental results indicate that the proposed method achieves the known minimum number of lines in many cases and has good scalability. Although the proposed method increases the quantum cost over a prior method based on functional decision diagrams, it significantly reduces the number of lines in most cases. Compared with the one-pass method using quantum multiple-valued decision diagrams, the proposed method reduces the quantum cost without increasing the number of lines in many cases. When compared with the lookup table-based method using a direct mapping flow, the method reduces the number of lines in a few cases. Thus, the method aids in the physical realization of a quantum circuit.

Locating the propagation source is one of the most important strategies to control the harmful diffusion process on complex networks. Most existing methods only consider the infection time information of the observers, but the diffusion direction information of the observers is ignored, which is helpful to locate the source. In this paper, we consider both of the diffusion direction information and the infection time information to locate the source. We introduce a relaxed direction-induced search (DIS) to utilize the diffusion direction information of the observers to approximate the actual diffusion tree on a network. Based on the relaxed DIS, we further utilize the infection time information of the observers to define two kinds of observers-based similarity measures, including the Infection Time Similarity and the Infection Time Order Similarity. With the two kinds of similarity measures and the relaxed DIS, a novel source locating method is proposed. We validate the performance of the proposed method on a series of synthetic and real networks. The experimental results show that the proposed method is feasible and effective in accurately locating the propagation source.