CSSP

CSSP estimates statistical power and required sequencing depth for ChIP-seq experiments using Bayesian estimation within a local Poisson model to guide experimental design.

Key Features:

• Statistical framework for power calculation: Calculates statistical power for ChIP-seq experiments across varying sequencing depths to estimate detection sensitivity.
• Local Poisson model: Models local read counts with a Poisson distribution consistent with methodologies adopted by many peak callers.
• Bayesian estimation: Applies Bayesian estimation techniques within the local Poisson framework for parameter inference.
• Simulation and data-driven validation: Validates power estimates using simulations and data-driven computational experiments based on pilot or preliminary data.
• Analytical sequencing depth determination: Provides analytical calculations to determine the sequencing depth required to achieve target power while controlling the false discovery rate (FDR).
• Utilization of existing data: Incorporates user-provided or publicly available ChIP-seq datasets to inform power and sequencing depth estimates.

Scientific Applications:

• Experiment design optimization: Guides selection of sequencing depth to optimize ChIP-seq experimental design and allocation of sequencing resources.
• Power analysis for fold changes: Assesses the ability to detect fold changes in ChIP enrichment over control samples, noting typical experiments are well-powered for large fold changes but often underpowered for smaller changes.

Methodology:

Bayesian estimation within a local Poisson model; simulations and data-driven computational experiments for validation; analytical calculations to compute required sequencing depth while controlling FDR.