Discovering functional impacts of miRNAs in cancers using a causal deep learning model
Related Experiment Video
Updated: Jan 31, 2026
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
Published on: October 27, 2023
Lujia Chen1, Xinghua Lu2,3,4
1Department of Biomedical Informatics, School of Medicine, University of Pittsburgh, 5607 Baum Blvd, Pittsburgh, PA, USA. luc17@pitt.edu.
Background:
Micro-RNAs (miRNAs) play a significant role in regulating gene expression under physiological and pathological conditions such as cancers. However, it remains a challenging problem to discover the target messenger RNAs (mRNAs) of a miRNA in a data driven fashion. On one hand, sequence-based methods for predicting miRNA targets tend to make too many false positive calls. On the other hand, analyzing expression correlation between miRNAs and mRNAs cannot establish whether relationship between a pair of correlated miRNA and mRNA is causal.
Methods:
In this study, we designed a deep learning model, referred to as miRNA causal deep net (mCADET), which aims to explicitly represent two types of statistical relationships between miRNAs and mRNAs: correlation resulting from confounded co-regulation and correlation as a result of causal regulation. The model utilizes a deep neural network to simulate transcription mechanism that leads to co-expression of miRNA and mRNA, and, in addition, it also contains directed edges from miRNAs to mRNAs to capture causal relationships among them.
Results:
We trained the mCADET model using pan-cancer miRNA and mRNA data from The Cancer Genome Atlas (TCGA) project to investigate mechanism of co-expression and causal interactions between miRNAs and mRNAs. Quantitative analyses of the results indicate that the mCADET significantly outperforms conventional deep learning models when modeling combined miRNA and mRNA expression data, indicating its superior capability of capturing the high-order statistical structures in the data. Qualitative analysis of predicted targets of miRNAs indicate that predictions by mCADET agree well with existing knowledge. Finally, the predictions by mCADET have a significantly lower false discovery rate and better overall accuracy in comparison to sequence-based and correlation-based methods when comparing to experimental results.
Conclusion:
The mCADET model can simultaneously infer the states of cellular signaling system regulating co-expression of miRNAs and mRNAs, while capturing their causal relationships in a data-driven fashion.
Insights
A new deep learning model, mCADET, accurately identifies causal relationships between microRNAs (miRNAs) and messenger RNAs (mRNAs) in cancer. This method improves upon existing techniques for predicting miRNA targets, reducing false positives and enhancing accuracy.
Area of Science:
- Genomics
- Bioinformatics
- Computational Biology
Background:
- Micro-RNAs (miRNAs) are crucial regulators of gene expression in both normal and cancerous conditions.
- Identifying miRNA targets (messenger RNAs or mRNAs) is challenging due to limitations of sequence-based and correlation-based prediction methods.
- Existing methods often yield high false positive rates or fail to establish causal links between miRNAs and mRNAs.
Purpose of the Study:
- To develop a novel deep learning model, miRNA causal deep net (mCADET), for accurate identification of miRNA-mRNA causal relationships.
- To explicitly model both co-regulation and causal regulation between miRNAs and mRNAs.
- To improve data-driven discovery of miRNA targets in cancer.
Main Methods:
- Designed mCADET, a deep neural network incorporating directed edges to capture causal miRNA-mRNA interactions.
- Simulated transcription mechanisms to model miRNA-mRNA co-expression.
- Trained the mCADET model using pan-cancer miRNA and mRNA expression data from The Cancer Genome Atlas (TCGA).
Main Results:
- mCADET demonstrated superior performance over conventional deep learning models in analyzing combined miRNA and mRNA expression data.
- The model effectively captured high-order statistical structures and outperformed existing methods in identifying miRNA targets.
- Predictions showed good agreement with existing biological knowledge and exhibited a significantly lower false discovery rate compared to sequence-based and correlation-based approaches.
Conclusions:
- The mCADET model offers a robust, data-driven approach for simultaneously inferring regulatory states and causal relationships between miRNAs and mRNAs.
- This advancement aids in understanding gene regulation in cancer and other complex biological systems.
- mCADET provides a more accurate and reliable tool for miRNA target prediction.
Related Concept Videos
01:21Causality in Epidemiology
01:26Higher Mental Functions of Brain: Learning and Memory
01:19Impact of Groups on Groups
01:28Criteria for Causality: Bradford Hill Criteria - II
01:30Criteria for Causality: Bradford Hill Criteria - I
01:30Impact
When particles with different initial velocities collide, they induce deformation by applying equal and opposite impulses. At the point of maximum deformation, the particles move together with...