PolyASite

PolyASite catalogs polyadenylation sites from 3' end sequencing data across human, mouse, and worm genomes using publicly available datasets from the Sequence Read Archive (SRA) curated through June 2019, enabling definition of alternative polyadenylation events and analysis of RNA 3' end cleavage and polyadenylation-mediated transcript isoform variation.

Key Features:

• Data source: Aggregates publicly available 3' end sequencing datasets from the Sequence Read Archive (SRA) curated through June 2019.
• Species coverage: Includes polyadenylation sites from human, mouse, and worm genomes.
• Identification of APA sites: Identifies and catalogs alternative polyadenylation (APA) sites derived from RNA 3' end sequencing data.
• Regulatory signal analysis: Detects signals related to RNA 3' end cleavage and polyadenylation and supports analysis of cell type-specific isoforms.
• Quality control: Integrates datasets using uniform quality measures and flags potential internal priming sites that could confound results.
• Clustering of sites: Clusters closely spaced polyadenylation sites that share similar signals to account for stochastic variation in processing.
• Representative site selection: Determines a representative site per cluster as the most frequently processed position.
• Quantification: Estimates relative use of representative sites across the transcriptome from all available samples.

Scientific Applications:

• Alternative polyadenylation profiling: Characterizes alternative polyadenylation patterns and predominant versus alternative cleavage sites.
• Transcript isoform analysis: Supports analysis of transcript isoform variation arising from alternative 3' end processing.
• Regulatory signal discovery: Enables identification of sequence or processing signals associated with RNA 3' end cleavage and polyadenylation.
• Cell type–specific studies: Facilitates detection of cell type-specific polyadenylation isoforms.
• Comparative analyses: Allows cross-species comparison of polyadenylation landscapes in model organisms (human, mouse, worm).

Methodology:

Aggregates 3' end sequencing datasets from SRA (curated through June 2019), applies uniform quality measures, identifies and flags potential internal priming sites, clusters closely spaced sites sharing similar signals, selects the most frequently processed site as a cluster representative, and estimates relative site usage across available samples.