stageR

stageR implements a two-stage testing procedure for stage-wise analysis of high-throughput gene and transcript-level expression data, controlling gene-level false discovery rate (FDR) while enabling post hoc inference on individual hypotheses.

Key Features:

• Two-Stage Testing Paradigm: Implements the stage-wise two-stage testing procedure (stageR) that aggregates multiple per-gene hypotheses into a screening stage and a follow-up confirmation stage.
• Enhanced Power and FDR Control: Increases statistical power, including for interaction effects, while controlling gene-level FDR.
• Transcript-Level Analysis Framework: Performs powerful aggregated gene-level tests combined with post hoc transcript-level assessments to preserve transcript-resolution interpretation.
• Applicability Across Complex Designs: Applicable to complex high-throughput experimental designs whenever individual hypotheses can be aggregated per gene.

Scientific Applications:

• Differential Expression Testing: Gene-level and transcript-level differential expression analysis in high-throughput expression studies.
• Interaction Effect Testing: Detection and improved power for interaction effects in multifactorial designs.
• Post Hoc Transcript/ Isoform Identification: Identification of significant transcripts or isoforms within genes following gene-level discovery.
• Analyses with Multiple Hypotheses per Gene: Studies requiring aggregation and joint control of multiple hypotheses mapped to the same gene.

Methodology:

Applies a two-stage testing paradigm: aggregated gene-level screening followed by post hoc confirmation of individual hypotheses (transcripts), aggregates multiple hypotheses per gene, and controls gene-level FDR while boosting power for interaction effects.