College of Computer and Control Engineering, Qiqihar University, Qiqihar, 161006, People's Republic of China.
Sci Rep. 2022 Oct 11;12(1):17030. doi: 10.1038/s41598-022-21243-y.
Non-coding RNAs (ncRNAs) have long been considered the "white elephant" on the genome because they lack the ability to encode proteins. However, in recent years, more and more biological experiments and clinical reports have proved that ncRNAs account for a large proportion in organisms. At the same time, they play a decisive role in the biological processes such as gene expression and cell growth and development. Recently, it has been found that short sequence non-coding RNA(miRNA) and long sequence non-coding RNA(lncRNA) can regulate each other, which plays an important role in various complex human diseases. In this paper, we used a new method (JSCSNCP-LMA) to predict lncRNA-miRNA with unknown associations. This method combined Jaccard similarity algorithm, self-tuning spectral clustering similarity algorithm, cosine similarity algorithm and known lncRNA-miRNA association networks, and used the consistency projection to complete the final prediction. The results showed that the AUC values of JSCSNCP-LMA in fivefold cross validation (fivefold CV) and leave-one-out cross validation (LOOCV) were 0.9145 and 0.9268, respectively. Compared with other models, we have successfully proved its superiority and good extensibility. Meanwhile, the model also used three different lncRNA-miRNA datasets in the fivefold CV experiment and obtained good results with AUC values of 0.9145, 0.9662 and 0.9505, respectively. Therefore, JSCSNCP-LMA will help to predict the associations between lncRNA and miRNA.
非编码 RNA（ncRNAs）长期以来一直被认为是基因组中的“白象”，因为它们缺乏编码蛋白质的能力。然而，近年来，越来越多的生物学实验和临床报告证明，ncRNAs 在生物体中占有很大比例。同时，它们在基因表达和细胞生长发育等生物过程中起着决定性的作用。最近，人们发现短序列非编码 RNA（miRNA）和长序列非编码 RNA（lncRNA）可以相互调节，这在各种复杂的人类疾病中起着重要作用。在本文中，我们使用了一种新的方法（JSCSNCP-LMA）来预测具有未知关联的 lncRNA-miRNA。该方法结合了 Jaccard 相似性算法、自调谐谱聚类相似性算法、余弦相似性算法和已知的 lncRNA-miRNA 关联网络，并使用一致性投影来完成最终的预测。结果表明，JSCSNCP-LMA 在五重交叉验证（fivefold CV）和留一交叉验证（LOOCV）中的 AUC 值分别为 0.9145 和 0.9268。与其他模型相比，我们成功地证明了其优越性和良好的可扩展性。同时，该模型还在五重 CV 实验中使用了三个不同的 lncRNA-miRNA 数据集，并分别获得了 AUC 值为 0.9145、0.9662 和 0.9505 的良好结果。因此，JSCSNCP-LMA 将有助于预测 lncRNA 与 miRNA 之间的关联。
