REA

REa: Statistical Analysis of RIP-chip Data

REa analyzes RNA immunoprecipitation combined with microarray (RIP-chip) datasets by correcting immunoprecipitation bias, removing background signal from non-target genes, and modeling measurement distributions to compute false discovery rates (FDRs).

Key Features:

• Background Subtraction: Removes non-target gene signal from normalized RIP-chip expression values to improve accuracy.
• Normalization for IP Efficiency Variability: Applies a secondary normalization step to correct biases introduced by variable immunoprecipitation (IP) efficiency across experiments and replicates.
• Gaussian Mixture Modeling: Models RIP-chip measurement distributions as Gaussian mixture distributions to estimate FDRs for user-defined cut-offs.
• Principal Component Analysis (PCA): Determines normalization factors to correct immunoprecipitation bias and enable cross-condition comparability.
• Validation with HITS-CLIP Data: Validated against published HITS-CLIP (High-Throughput Sequencing of RNA isolated by CrossLinking ImmunoPrecipitation) datasets generated from the same cell line.

Scientific Applications:

• mRNA Target Identification: Identifies mRNA targets of RNA-binding proteins, including Ago2, a core component of the microRNA-guided RNA-induced silencing complex (RISC), to study post-transcriptional gene regulation.

Methodology:

Processes normalized RIP-chip expression values, performs background subtraction to remove non-target genes, applies principal component analysis to correct IP efficiency variability, and fits Gaussian mixture models to define enriched transcripts and compute false discovery rates.