Multilocus Selection Measurement: Genomic Simulations for p ≫ n Scenarios

Zachariah Gompert produced this simulation dataset in 2020 to test multilocus approaches for measuring selection on correlated genetic loci. It models high-dimensional genomic data where the number of parameters exceeds observations (p ≫ n) to evaluate the efficacy of Bayesian sparse linear mixed models.

Use Cases

• Quantifying direct selection gradients using genomic variation and fitness associations
• Benchmarking Bayesian sparse linear mixed models for genome-wide association mapping
• Simulating ecological speciation by analyzing correlations between trait values and survival

Strengths

• Employs Bayesian sparse linear mixed models (BSLMM) for high-dimensional data
• Released under a CC0-1.0 public domain license
• Addresses the p ≫ n dimensionality challenge in evolutionary biology

Limitations

• Data is entirely synthetic/simulated rather than empirical
• Metadata does not specify exact row counts or column headers
• Temporal staleness with the last update occurring in 2020

Provenance

Source

Dryad Digital Repository

Collection Method

synthetic

Freshness

Last updated June 2020; no scheduled updates.