MAFsnp

MAFsnp employs a frequentist statistical framework to identify single nucleotide polymorphisms (SNPs) from next-generation sequencing (NGS) data across multiple samples for accurate variant detection.

Key Features:

• Multiple-sample NGS SNP calling: Calls SNPs using data aggregated across multiple samples from next-generation sequencing.
• Frequentist eLRT statistic: Uses an estimated likelihood ratio test (eLRT) statistic within a frequentist framework as an alternative to Bayesian statistical frameworks.
• Two-parameter mixture distribution modeling: Models the eLRT statistic with a two-parameter mixture distribution to address parameters near the boundary of the parametric space.
• P-value calculation: Produces p-values directly from the modeled test statistic for SNP detection.
• FDR control via multiple-testing correction: Supports multiple-testing correction to control the false discovery rate (FDR) at pre-specified levels.
• Robustness to finite-sample issues: Accounts for inadequacy of standard large-sample properties in finite-sample distributions when parameters lie near boundaries.
• Demonstrated accuracy: Comparative analyses on simulated and real genomic datasets indicate improved SNP-calling accuracy and reliability relative to existing callers.

Scientific Applications:

• High-throughput genetic studies: Suitable for large-scale variant discovery where stringent FDR control is required.
• Complex disease genetics: Applicable to studies seeking accurate SNP detection to associate variants with disease phenotypes.
• Population genetics: Enables reliable identification of population-level SNP variation from NGS data.
• Evolutionary biology: Facilitates detection of nucleotide polymorphisms relevant to evolutionary analyses.

Methodology:

Computational methods explicitly include an estimated likelihood ratio test (eLRT) statistic, modeling the eLRT with a two-parameter mixture distribution to obtain p-values, and applying multiple-testing correction to control the false discovery rate (FDR).