microTSS

microTSS predicts transcription start sites (TSSs) at single-nucleotide resolution for intergenic microRNAs (miRNAs) using a machine-learning approach that integrates high-resolution RNA-sequencing, chromatin immunoprecipitation (ChIP) transcription marks, and DNase-seq.

Key Features:

• Integration of High-Resolution Data: Integrates high-resolution RNA-sequencing with active transcription marks from chromatin immunoprecipitation (ChIP) and DNase-seq to enhance TSS detection.
• Single-Nucleotide Resolution: Provides TSS predictions at single-nucleotide resolution specifically for intergenic miRNAs.
• Tissue-Specific Characterization: Incorporates data from multiple tissues to characterize tissue-specific promoters for intergenic miRNAs.
• Validation and Applicability: Predictions have been validated using a Drosha-null/conditional-null mouse model generated via the conditional by inversion (COIN) methodology.
• Integration into Regulatory Network Modelling: Produces miRNA TSS annotations suitable for incorporation of miRNA transcription regulation into tissue-specific regulatory network models.

Scientific Applications:

• Promoter Identification: Identifies promoters for intergenic miRNAs to support studies of miRNA gene regulation.
• Regulatory Network Modelling: Contributes miRNA transcription initiation data for constructing tissue-specific regulatory network models.
• Comparative Genomics: Supports comparative analyses of human and mouse pri-miRNAs, including detection of divergent transcription at active gene promoters and overlaps with long non-coding RNAs.

Methodology:

Implements a machine-learning algorithm that integrates high-resolution RNA-sequencing, chromatin immunoprecipitation (ChIP) transcription marks, and DNase-seq to predict TSSs at single-nucleotide resolution.