miRStart 2.0

miRStart 2.0 annotates miRNA transcription start sites (TSSs) and identifies tissue-specific transcription factor (TF)-miRNA interactions using a deep learning-based model integrated with over 4500 high-throughput datasets.

Key Features:

• Deep learning-based TSS prediction: A multi-modal deep learning model predicts miRNA TSSs, achieving the most overlaps with both cell-specific and non-cell-specific validated TSSs compared to existing tools.
• Multi-omics data integration: The tool integrates over 4500 high-throughput datasets across five data types, including ChIP-seq, DNase hypersensitivity, and UCSC conservation data.
• Comprehensive TSS annotation: miRStart 2.0 annotates 28,828 putative TSSs for 1,745 human and 1,181 mouse miRNAs, supported by sequencing-based signals.
• TF-miRNA interaction mapping: The database contains over 6 million tissue-specific TF-miRNA interactions derived from ChIP-seq data.
• Network visualization: Network visualizations are provided to facilitate miRNA regulatory insights.

Scientific Applications:

• miRNA transcriptional regulation research: miRStart 2.0 enables identification of miRNA upstream regulatory elements in relation to their TSSs.
• Systems-level gene regulation analysis: The tool provides insights into gene regulation and disease mechanisms at a systems level.
• Therapeutic target discovery: miRStart 2.0 facilitates the discovery of novel therapeutic targets and precision medicine strategies through translational research.

Methodology:

miRStart 2.0 utilizes a multi-modal deep learning-based model integrated with over 4500 high-throughput datasets across five data types, incorporating ChIP-seq data, DNase hypersensitivity data, and UCSC conservation data.