MDA-GCNFTG: identifying miRNA-disease associations based on graph convolutional networks via graph sampling through the feature and topology graph

Abstract

Accurate identification of the miRNA-disease associations (MDAs) helps to understand the etiology and mechanisms of various diseases. However, the experimental methods are costly and time-consuming. Thus, it is urgent to develop computational methods towards the prediction of MDAs. Based on the graph theory, the MDA prediction is regarded as a node classification task in the present study. To solve this task, we propose a novel method MDA-GCNFTG, which predicts MDAs based on Graph Convolutional Networks (GCNs) via graph sampling through the Feature and Topology Graph to improve the training efficiency and accuracy. This method models both the potential connections of feature space and the structural relationships of MDA data. The nodes of the graphs are represented by the disease semantic similarity, miRNA functional similarity and Gaussian interaction profile kernel similarity. Moreover, we considered six tasks simultaneously on the MDA prediction problem at the first time, which ensure that under both balanced and unbalanced sample distribution, MDA-GCNFTG can predict not only new MDAs but also new diseases without known related miRNAs and new miRNAs without known related diseases. The results of 5-fold cross-validation show that the MDA-GCNFTG method has achieved satisfactory performance on all six tasks and is significantly superior to the classic machine learning methods and the state-of-the-art MDA prediction methods. Moreover, the effectiveness of GCNs via the graph sampling strategy and the feature and topology graph in MDA-GCNFTG has also been demonstrated. More importantly, case studies for two diseases and three miRNAs are conducted and achieved satisfactory performance.

Predicting the hosts of prokaryotic viruses using GCN-based semi-supervised learning

“…In biological data analyses, there exist many topological structures such as gene-sharing network, disease-drug relationship graph, and diseases-gene relationship graph. Utilizing these relationships in GCN has led to several successful applications [21,22,23,24,25].…”

Multiscale molecular modelling: from electronic structure to dynamics of nanosystems and beyond

“…So, I will be content with a few general reviews and some papers that I found particularly interesting or promising. The focus will be on methods that retain a structural/dynamic model where the role of ML is to accelerate calculations/simulations rather than the, equally meritorious, approach of data base searches, for example in the area of drug design (where we have made some recent contributions [126][127][128] ). Below, I will follow the bi-thematic approach adopted so far in this Tutorial Review with sights firmly set on complex materials and biological systems.…”

BLNIMDA: identifying miRNA-disease associations based on weighted bi-level network

“…The final association prediction was equated to the recommendation task, and a matrix completion was applied to predict hidden MDAs. For predicting MDAs from a comprehensive and novel perspective, Chu et al [ 17 ] developed an original model (MDA-DCNFTG) based on the GCN, which treated MDAs prediction as a node classification task. The highlight of the MDA-GCNFTG was that it used graph sampling to predict MDAs from the perspective of feature and topological graphs based on Graph Convolutional Networks (GCNs).…”