SweeD

SweeD detects selective sweeps in genomic SNP (Single Nucleotide Polymorphism) data and estimates their locations and selection coefficients while accounting for ascertainment biases and complex demographic models.

Key Features:

• Composite Likelihood-Based Parametric Test: Implements a composite likelihood–based parametric test built on the SweepFinder algorithm to detect selective sweeps with high power and low Type I error under varying recombination rates and demography.
• Robustness to Demographic Assumptions: Maintains detection accuracy across complex demographic scenarios and varying mutation and recombination rates.
• Correction for Ascertainment Biases: Corrects for ascertainment biases inherent in SNP discovery processes to prevent skewed inference.
• Estimation of Sweep Parameters: Estimates sweep location and the magnitude of the selection coefficient for detected sweeps.
• Theoretical SFS Calculation: Computes the theoretical Site Frequency Spectrum (SFS) for specified demographic models following Zickovic and Stephan (2011), including stepwise change and exponential growth followed by stepwise change scenarios.

Scientific Applications:

• Identifying Regions Under Positive Selection: Pinpoints genomic regions that experienced recent positive selection, exemplified by analyses of the lactase gene on Chromosome 2.
• Exploring Disease Associations: Detects selective sweeps in genes associated with disease risk such as DPP10 and COL4A3 to inform evolutionary interpretations of disease-related loci.
• Analyzing Population Genetics Data: Analyzes SNP datasets from projects such as the Seattle SNP project and the HapMap project across populations with complex histories.

Methodology:

Implements a composite likelihood–based parametric test built on the SweepFinder algorithm, applies corrections for ascertainment bias, estimates sweep location and selection coefficient, and computes theoretical SFS according to Zickovic and Stephan (2011).