Download PDFOpen PDF in browser

Predicting the Human miRNA-Disease Associations Based on Non-Linear Gaussian Profile Kernel Similarity

EasyChair Preprint no. 9377

10 pagesDate: November 28, 2022

Abstract

In view of the fact that the traditional methods of determining potential miRNA-disease associations tend to be destructive, labor-intensive, time-consuming, and associated with practice effects, a number of computational methods are being developed to address the burden on biological researchers. In this study, we introduced the computational strategy of non-linear gaussian profile kernel similarity and proposed a novel deep-learning method called NGPKS to engage the in-depth understanding of miRNA-disease associations. More specifically, NGPKS comprehensively integrates the miRNA functional similarity and disease semantic similarity information. Then, the gaussian interaction profile kernel similarity algorithm was utilized to capture the structural information between miRNAs and diseases. Finally, a deep learning framework was constructed for modeling the integration of two types of similarity features. We used three model validation strategies, including five-fold cross-validation, comparison with the state-of-the-art methods, and ablation experiments were used to check the predictive ability of our model. Besides, we conducted case studies for two common diseases. As a result, there are 50 (Colon Cancer), and 47(Lymphoma) among the top 50 predicted miRNAs validated through experiments. Therefore, we could conclude that NGPKS is an effective method to predict potential miRNA-disease associations.

Keyphrases: machine learning, miRNA-disease matrix, NGPKS

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@Booklet{EasyChair:9377,
  author = {Zou Haitao and Xie Xiaolan and Ji Boya and Peng Shaoliang},
  title = {Predicting the Human miRNA-Disease Associations Based on Non-Linear Gaussian Profile Kernel Similarity},
  howpublished = {EasyChair Preprint no. 9377},

  year = {EasyChair, 2022}}
Download PDFOpen PDF in browser