PRAM

PRAM assembles transcript models from pooled RNA-seq datasets to discover intergenic and unannotated transcripts and associated splice junctions.

Key Features:

• Innovative Methodology: Pools multiple RNA-seq datasets and assembles transcript models in a single '1-Step' approach rather than building transcripts per-sample then merging ('2-Step').
• Enhanced Prediction Accuracy: Demonstrates superior performance in predicting overall transcript structures and individual splice junctions and is competitive in detecting exonic nucleotides based on computational benchmarks.
• Application to Large Datasets: Designed for use with large collections of RNA-seq datasets and retrospective analyses of public RNA-seq data.
• Discovery of Unannotated Transcripts: Identified unannotated transcripts in 30 human ENCODE RNA-seq datasets that exhibit epigenetic and RAMPAGE signatures similar to recently annotated transcripts.
• Experimental Validation: Facilitated discovery and experimental validation of novel mouse hematopoietic transcripts, including differentially expressed transcripts linked to Pik3cg.

Scientific Applications:

• Transcriptome Exploration: Identification and annotation of novel intergenic and unannotated transcripts from large-scale RNA-seq data.
• Epigenetic Studies: Detection of transcripts with epigenetic and RAMPAGE signatures to investigate gene regulatory mechanisms.
• Comparative Genomics: Discovery of conserved transcriptomic relationships across human and mouse datasets, exemplified by transcripts associated with Pik3cg.

Methodology:

Pools multiple RNA-seq datasets and assembles transcript models in a single '1-Step' approach, benchmarks against '2-Step' methods, and compares predicted transcripts to epigenetic and RAMPAGE signatures; applied to 30 human ENCODE and mouse hematopoietic RNA-seq datasets.