MicroRNAs (miRNA) are a type of non-coding RNA molecules that are effective on the formation and the progression of many different diseases. Various researches have reported that miRNAs play a major role in the prevention, diagnosis, and treatment of complex human diseases. In recent years, researchers have made a tremendous effort to find the potential relationships between miRNAs and diseases. Since the experimental techniques used to find that new miRNA-disease relationships are time-consuming and expensive, many computational techniques have been developed. In this study, Weighted [Formula: see text]-Nearest Known Neighbors and Network Consistency Projection techniques were suggested to predict new miRNA-disease relationships using various types of knowledge such as known miRNA-disease relationships, functional similarity of miRNA, and disease semantic similarity. An average AUC of 0.9037 and 0.9168 were calculated in our method by 5-fold and leave-one-out cross validation, respectively. Case studies of breast, lung, and colon neoplasms were applied to prove the performance of our proposed technique, and the results confirmed the predictive reliability of this method. Therefore, reported experimental results have shown that our proposed method can be used as a reliable computational model to reveal potential relationships between miRNAs and diseases.

中文翻译：

---

基于加权 K-最近已知邻居和网络一致性投影的 miRNA 疾病关联预测

微小RNA（miRNA）是一类非编码RNA分子，对许多不同疾病的形成和进展有效。各种研究报道，miRNA在复杂人类疾病的预防、诊断和治疗中发挥着重要作用。近年来，研究人员为寻找 miRNA 与疾病之间的潜在关系付出了巨大的努力。由于用于发现新的 miRNA 与疾病关系的实验技术既耗时又昂贵，因此开发了许多计算技术。在这项研究中，建议使用加权 [公式：见正文]-最近已知邻居和网络一致性投影技术来预测新的 miRNA-疾病关系，使用各种类型的知识，例如已知的 miRNA-疾病关系、miRNA 的功能相似性、和疾病语义相似性。在我们的方法中，分别通过 5 倍和留一法交叉验证计算了 0.9037 和 0.9168 的平均 AUC。应用乳腺、肺和结肠肿瘤的案例研究来证明我们提出的技术的性能，结果证实了该方法的预测可靠性。因此，报告的实验结果表明，我们提出的方法可以作为一种可靠的计算模型来揭示 miRNA 与疾病之间的潜在关系。