Datasets:
license: cc-by-4.0
tags:
- biology
- rna
- genomics
Overview
miRNA target prediction annotates which microRNAs (miRNAs) target a given mRNA transcript. This task is framed as a set of binary classification problems, one per miRNA, where an mRNA may be targeted by multiple miRNAs. The dataset is derived from experimentally validated miRNA–mRNA interactions identified using chimeric read-based assays.
This dataset is redistributed as part of mRNABench: https://github.com/morrislab/mRNABench
Data Format
Description of data columns:
target_{miRNA}: Binary indicator (0/1) denoting whether the corresponding miRNA targets the mRNA. The task includes the 20 most frequently observed human miRNAs in the source dataset.cds: Binary track which reports position of first nucleotide in each codon in the coding sequence (CDS).splice: Binary track which reports position of the 3' end of each exon, indicating splice sites.
Each row corresponds to a single representative transcript per gene, selected from top-priority transcripts. See the dataset_utils.py file under mrna_bench/datasets on the github for more information on how 'priority' is determined.
Data Source
This dataset was generated from the mirTarCLASH database, which is built from experiments that capture and sequence miRNA–mRNA hybrid molecules, directly identifying targeting interactions. Interaction reliability is supported by read depth and mutation signatures at binding sites. To mitigate severe class imbalance, the benchmark focuses on the 20 most prevalent miRNAs, and includes only transcripts targeted by at least one of them.
The dataset is redistributed under the CC BY 4.0 license. Please provide proper attribution if you use this dataset.
Please attribute:
Original paper: https://academic.oup.com/database/article/2025/baaf023/8090947
Original dataset source: https://cosbi.ee.ncku.edu.tw/MirTarClash/home/
Citation: Yang T-H et al., “mirTarCLASH: a comprehensive miRNA target database based on chimeric read-based experiments.” Database, 2025. https://doi.org/10.1093/database/baaf023