Statistical Methods

This section provides detailed documentation of the statistical methods, algorithms, and biological rationale underlying WASP2’s allele-specific analysis pipeline.

Contents

• Allele Counting Algorithm
  • Overview
  • Biological Rationale
  • Algorithm Details
  • Rust Acceleration
  • Quality Filtering
  • Output Format
  • Region Assignment
  • See Also
• WASP Mapping Bias Correction
  • The Problem: Reference Mapping Bias
  • The WASP Algorithm
  • Implementation Details
  • Rust Acceleration
  • Performance Characteristics
  • Limitations and Considerations
  • Pipeline Integration
  • See Also
• Beta-Binomial Model for Allelic Imbalance
  • Motivation
  • The Beta-Binomial Model
  • Hypothesis Testing
  • Implementation in WASP2
  • Phased vs Unphased Analysis
  • Output Interpretation
  • Practical Considerations
  • See Also
  • References
• Dispersion Parameter Estimation
  • The Dispersion Parameter
  • Estimation Methods
  • Maximum Likelihood Estimation
  • Method of Moments
  • Comparison: MLE vs MoM
  • Practical Considerations
  • See Also
  • References
• False Discovery Rate Correction
  • The Multiple Testing Problem
  • Error Rate Definitions
  • Benjamini-Hochberg Procedure
  • Alternative Methods
  • Alternative Correction Methods
  • Practical Guidelines
  • Output Format
  • References

Overview

WASP2 implements a complete pipeline for allele-specific analysis:

1. Allele Counting: Reads are assigned to reference or alternate alleles at heterozygous variant sites using base-level alignment information.

2. Mapping Bias Correction: The WASP algorithm removes reads that exhibit mapping bias by testing whether allele-swapped reads map to the same location.

3. Statistical Testing: Beta-binomial models account for overdispersion in allele count data when testing for allelic imbalance.

4. Multiple Testing Correction: False discovery rate control ensures reliable detection of true imbalanced regions.

References

[vandeGeijn2015]

van de Geijn B, McVicker G, Gilad Y, Pritchard JK (2015). WASP: allele-specific software for robust molecular quantitative trait locus discovery. Nature Methods 12:1061-1063.

[Castel2015]

Castel SE, Levy-Moonshine A, Mohammadi P, Banks E, Lappalainen T (2015). Tools and best practices for data processing in allelic expression analysis. Genome Biology 16:195.

[Skelly2011]

Skelly DA, Johansson M, Madeoy J, Wakefield J, Akey JM (2011). A powerful and flexible statistical framework for testing hypotheses of allele-specific gene expression from RNA-seq data. Genome Research 21:1728-1737.