#29 · Genomics
Sparse, mixed-source aDNA sampling limits kinship and selection inference
The dominant thread is that sampling design, rather than estimator choice alone, determines whether population-genetic inference is valid. A simulation study of sedimentary aDNA finds that multi-contributor samples impair kinship inference and proposes a heterozygosity test to flag them. A benchmark of four allele-frequency time-series selection methods (ApproxWF, BMWS, Slattice, Sr) on simulated aDNA-like sparse data assesses their power under these conditions. In domestication genomics, a modified STRUCTURE parameterization reduces sampling and Ne biases in the walnut analysis, supporting a South Asian origin. The amphioxus mutation-rate study addresses a different question and is adjacent only. What is new is treating sparse and mixed-source sampling as a determinant of inference validity rather than background noise.
Published papers
Assessing Ancient DNA Sampling Strategies for Natural Selection Inference in Humans Using Allele Frequency Time Series Data.
Benchmarks four allele-frequency time-series selection inference methods (ApproxWF, BMWS, Slattice, Sr) on simulated aDNA-like sparse datasets, assessing power…
Estimating the Reduction in Genetic Diversity from Background Selection under Non-equilibrium Demography and Partial Selfing.
Tests whether background selection reduction in diversity can be estimated under demographic change and partial selfing; proposes joint inference from windowed…
Resolving sampling and population-size biases in domestication genomics supports a South Asian origin of walnuts.
Proposes a modified STRUCTURE parameterization (F model, alternative ancestry prior, small ALPHA) that reduces sampling and Ne biases, applied to walnut…
Topic history
| Date | Rank |
|---|---|
| #29 | |
| #26 |