PRAM

The software tool "PRAM" (Pooling RNA-seq and Assembling Models) introduces a 1-Step transcript discovery approach in extensive RNA-seq dataset collections. Unlike current 2-step methods that build transcripts from individual datasets and then merge predictions, PRAM directly constructs transcript models from pooled RNA-seq datasets to enhance the power of transcript discovery.
In a computational benchmark, PRAM's 1-step approach outperforms 2-step methods in predicting overall transcript structures and individual splice junctions while remaining competitive in detecting exonic nucleotides. Applying PRAM to 30 human ENCODE RNA-seq datasets reveals unannotated transcripts with signatures similar to recently annotated ones. PRAM identifies and experimentally validates new transcripts in a case study using mouse hematopoietic RNA-seq datasets. These transcripts exhibit a differential expression pattern with a neighboring gene, Pik3cg, implicated in human hematopoietic phenotypes. The conservation of this relationship in humans is supported by evidence.