miRFANs: an integrated database for Arabidopsis thalianamicroRNA function annotations

Hui Liu, Ting Jin, Ruiqi Liao, Linxia Wan, Bin Xu, Shuigeng Zhou, Jihong GuanBMC Plant Biology 2012 · DOI

Abstract

miRFANs is a comprehensive, integrated database designed to decode the functional roles of microRNAs (miRNAs) in the model plant Arabidopsis thaliana. By combining miRNA-target predictions with real-world expression data and transcription factor networks, it moves beyond simple "lists of targets" to provide a spatio-temporal map of how these small molecules regulate plant development and stress responses.

Introduction

MicroRNAs (miRNAs) are short RNA molecules that act as "dimmer switches" for genes, typically by binding to messenger RNAs (mRNAs) and preventing their translation. While crucial for processes ranging from hormone secretion to drought survival, two major hurdles persist in miRNA research: the high rate of false positives in computational predictions and the extreme spatio-temporal specificity of miRNA regulation. Existing databases often lack the functional annotations or comparative expression data needed to confirm if a miRNA and its predicted target are active in the same tissue at the same time.

System Architecture and Methods

The authors developed miRFANs to bridge these gaps by integrating multiple biological data layers and analysis tools. The system's architecture is divided into an integrative database and a user-friendly web interface.

System architecture of the miRFANs database, showing the integration of various data sources and the web interface modules.

The database aggregates miRNA sequences from miRBase, genomic data from TAIR, and extensive expression profiles from GEO, ASRP, and PMRD. To improve target prediction accuracy, miRFANs integrates three algorithms (psRNATarget, TargetAlign, and UEA target-finder) alongside manually curated experimental data from TarBase and miRTarBase.

A core innovation of miRFANs is its "secret sauce": differential expression analysis. The system compares expression profiles of miRNAs and their targets from identical or similar experimental conditions. Reciprocal expression patterns—where a miRNA increases while its target decreases—provide strong evidence of a functional regulatory event.

Workflow for the functional annotation of miRNAs, illustrating the process from sequence matching to condition-specific functional mapping.

Furthermore, the database integrates Transcription Factor (TF) data from the AGRIS database. This allows for the construction of miRNA-mediated genetic regulatory networks, acknowledging that gene expression is controlled by both "top-down" TF activation and "bottom-up" miRNA interference.

Key Results and Insights

At the time of publication, miRFANs processed expression data from 12 miRNA experiments (81 samples) and 548 genome-wide microarray experiments (6,740 samples). Key achievements include:

  • Comprehensive Target Mapping: Integration of 81 experimentally validated interactions and thousands of high-confidence predictions.
  • Regulatory Network Discovery: Identification of a network consisting of 12 miRNA families, 41 TFs, and 155 regulatory relationships.
  • Functional Context: Every target is linked to Gene Ontology (GO) terms and KEGG pathways, enabling researchers to identify specific conditions (e.g., salt stress) where miRNA-target pairs are active.

Discussion and Future Work

miRFANs provides a significant advancement over static target lists by offering a dynamic "mining toolbox" for correlation analysis and clustering. While currently limited to Arabidopsis thaliana, the authors plan to expand the platform to other model plants. Future iterations will focus on deeper integration into broader transcriptional regulatory networks to create a more holistic model of plant systems biology. Despite the power of multi-algorithmic integration, the authors emphasize that computational predictions still require final "wet lab" validation.

Summary updated

BibTeX
@inproceedings{liu2012mirfans,
  author = {Hui Liu and Ting Jin and Ruiqi Liao and Linxia Wan and Bin Xu and Shuigeng Zhou and Jihong Guan},
  title = {{miRFANs: an integrated database for Arabidopsis thalianamicroRNA function annotations}},
  booktitle = {BMC Plant Biology},
  year = {2012},
  doi = {10.1186/1471-2229-12-68},
}
Figure