Cepred: Predicting the Co-Expression Patterns of the Human Intronic microRNAs with Their Host Genes

Dong Wang, Ming Lü, Jing Miao, Tingting Li, Edwin Wang, Qinghua CuiPLoS 2009 · DOI

Introduction

MicroRNAs (miRNAs) are critical regulators of gene expression, influencing biological processes and disease states. A significant challenge in miRNA research is identifying the specific tissues where they are expressed. Direct detection methods are often limited by cost, time, and the low abundance of certain miRNAs.

Approximately 36% of human miRNAs are "intronic," residing within the non-coding regions (introns) of protein-coding host genes. While many intronic miRNAs are co-expressed with their hosts, others are regulated independently. This study introduces Cepred, a machine learning framework designed to predict co-expression patterns between intronic miRNAs and their host genes using simple genomic positional features.

Methods and Model Architecture

The researchers developed Cepred using a Support Vector Machine (SVM) classifier. The model's primary innovation is its reliance on "genomic geography"—the physical coordinates of the miRNA within its host gene—rather than complex biochemical signatures.

The SVM was trained using four specific positional features, as detailed in the table below:

Table 1: Description of the four positional features used as inputs for the Cepred model.

These features allow the model to distinguish between miRNAs that "hitchhike" on their host's transcriptional machinery and those with independent regulatory mechanisms. The training set consisted of 29 miRNA-host gene pairs with established co-expression levels.

Results and Validation

Cepred demonstrated high predictive power across multiple evaluation metrics:

  • Cross-Validation: The model achieved 79% accuracy in leave-one-out cross-validation.
  • Independent Testing: On a separate dataset of 21 pairs curated from multiple laboratories, the model reached 95% accuracy.
  • Tissue-Specific Inference: By applying Cepred to 178 intronic miRNAs, the authors inferred tissue-specific expression profiles based on host gene data from the Su et al. (2004) gene atlas. For instance, the model accurately predicted the nervous system expression of mir-488 and the cardiac/skeletal muscle expression of mir-208b.

Discussion

The success of Cepred suggests that the spatial arrangement of a miRNA within a gene is a dominant factor in its biogenesis. This "shortcut" allows researchers to use host gene expression (easily tracked via standard microarrays) as a proxy for miRNA activity.

While currently limited to intronic miRNAs and a binary classification (high vs. low co-expression), the framework provides a scalable foundation. Future iterations could incorporate regression models to predict exact correlation degrees or expand to intergenic miRNAs. Ultimately, Cepred offers a valuable computational tool for mapping the regulatory landscape of the human genome.

Summary updated

Abstract

Identifying the tissues in which a microRNA is expressed could enhance the understanding of the functions, the biological processes, and the diseases associated with that microRNA. However, the mechanisms of microRNA biogenesis and expression remain largely unclear and the identification of the tissues in which a microRNA is expressed is limited. Here, we present a machine learning based approach to predict whether an intronic microRNA show high co-expression with its host gene, by doing so, we could infer the tissues in which a microRNA is high expressed through the expression profile of its host gene. Our approach is able to achieve an accuracy of 79% in the leave-one-out cross validation and 95% on an independent testing dataset. We further estimated our method through comparing the predicted tissue specific microRNAs and the tissue specific microRNAs identified by biological experiments. This study presented a valuable tool to predict the co-expression patterns between human intronic microRNAs and their host genes, which would also help to understand the microRNA expression and regulation mechanisms. Finally, this framework can be easily extended to other species.
BibTeX
@inproceedings{wang2009cepred,
  author = {Dong Wang and Ming Lü and Jing Miao and Tingting Li and Edwin Wang and Qinghua Cui},
  title = {{Cepred: Predicting the Co-Expression Patterns of the Human Intronic microRNAs with Their Host Genes}},
  booktitle = {PLoS},
  year = {2009},
  doi = {10.1371/journal.pone.0004421},
}

Loading related content...

Figure