SSL: Inferring disease-related genes using sentence structure and literature data

Text mining is widely applied in biology to infer relationships between biological entities. In biology, disease-gene relationships are important to discover the cause of disease. Therefore, we propose a useful method called SSL, which infers disease-related genes, using sentence structure and literature data. Using sentence structure, the proposed method decreases the number of candidate disease-related genes and infers more meaningful disease-related genes than other comparable methods. Furthermore, our method extracts useful sentences that have information on the relationship between specific diseases and genes. By analyzing the structure of the sentences, we can obtain useful knowledge of disease-gene relationships. We applied our method to five diseases, including Alzheimer's disease, prostate cancer, gastric cancer, colorectal cancer, and lung cancer. For validation, we investigated the top 10 inferred genes for five diseases. Our method demonstrated up to 50% higher precision than existing methods, and showed 98% accuracy in inferring disease-related genes.