Abstract:Circular RNA (circRNA) is a kind of novel discovered non-coding RNA molecule with a closed loop structure, which plays a critical regulatory role in human diseases. Identifying the association between circRNAs and diseases has important potential value for the diagnosis and treatment of complex human diseases. Although biological experiments can more accurately identify the association between circRNAs and diseases, they are usually blind and limited by small scale and high cost. Therefore, there is an urgent need for efficient and feasible computational methods to predict the potential circRNA-disease associations on a large scale, so as to provide the most promising candidate for biological experiments. In this paper, we propose a novel computational method based on the deep Generative Adversarial Network (GAN) algorithm combined with the multi-source similarity information to predict the circRNA-disease associations. Firstly, we fuse the multi-source information of disease semantic similarity, disease and circRNA Gaussian interaction profile kernel similarity, and then use GAN to extract the hidden features of fusion information objectively and effectively in the way of confrontation learning, and finally send them to Logistic Model Tree (LMT) classifier for accurate prediction. The 5-fold cross-validation experiment of the proposed model achieved 89.2% accuracy with 89.4% precision at the AUC of 90.6% on the CIRCR2Disease dataset. Compared with the state-of-the-art SVM classifier and other feature extraction methods, the proposed model shows strong competitiveness. In addition, the predicted results of this model are supported by the biological experiments, and 9 of the top 15 circRNA-disease associations with the highest scores were confirmed by recently published literature. These promising results indicate that the proposed model is an effective tool for predicting circRNA-disease associations and can provide reliable candidates for biological experiments.

Rare Disease Detection by Sequence Modeling with Generative Adversarial Networks

Semi-supervised Rare Disease Detection Using Generative Adversarial Network

A Method for the Early Prediction of Chronic Diseases Based on Short Sequential Medical Data.

Generating Synthetic Mixed-Type Longitudinal Electronic Health Records for Artificial Intelligent Applications

A Generative Adversarial Network Model for Disease Gene Prediction With RNA-seq Data

Multi-Label Clinical Time-Series Generation via Conditional GAN

Few-shot Radiology Report Generation for Rare Diseases

Predicting LncRNA-Disease Association Based on Generative Adversarial Network

Predicting Circrna-Disease Associations Using Deep Generative Adversarial Network Based on Multi-Source Fusion Information.

A Semi-Supervised Generative Adversarial Network for Prediction of Genetic Disease Outcomes

GANCDA: a Novel Method for Predicting Circrna-Disease Associations Based on Deep Generative Adversarial Network

A Model-Based Reinforcement Learning Approach for a Rare Disease Diagnostic Task

Difficulty-aware Meta-learning for Rare Disease Diagnosis

Performance and clinical utility of a new supervised machine-learning pipeline in detecting rare ciliopathy patients based on deep phenotyping from electronic health records and semantic similarity

Electrocardiogram generation with a bidirectional LSTM-CNN generative adversarial network

Discriminative ensemble meta-learning with co-regularization for rare fundus diseases diagnosis

CONAN: Complementary Pattern Augmentation for Rare Disease Detection

A Bootstrap Machine Learning Approach to Identify Rare Disease Patients from Electronic Health Records

Assessing DxGPT: Diagnosing Rare Diseases with Various Large Language Models

Rare disease diagnosis using knowledge guided retrieval augmentation for ChatGPT