Predicting microRNA targets with a new model
Author Information
Author(s): Saito Takaya, Sætrom Pål
Primary Institution: Norwegian University of Science and Technology
Hypothesis
By including non-stringent seed sites and training two separate SVMs on target site and mRNA-level features, we would create an accurate miRNA target prediction method with high sensitivity and overall prediction performance.
Conclusion
The two-step SVM model outperformed existing algorithms in identifying true miRNA targets among genes with conserved target sites.
Supporting Evidence
- The two-step model showed higher performance than established algorithms.
- The model does not rely on conservation for predicting targets.
- It incorporates both individual target site characteristics and global mRNA properties.
Takeaway
The researchers created a new way to predict which genes are targeted by microRNAs, and it works better than older methods.
Methodology
The model uses support vector machines (SVM) and is trained on large-scale microarray datasets.
Potential Biases
Potential bias from using data from human cell lines may affect the model's performance in other species.
Limitations
The model is trained on data from over-expression experiments, which may introduce bias from interactions with endogenous miRNAs.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website