Table 1.
Examples of data types and models that can aid the quantification of uncertainty related to predicting evolution over moderate time scales.
| Data type | Model | Key features | Software (citation) |
|---|---|---|---|
| Trait genetics | Bayesian sparse linear mixed model (BSLMM) | Estimates heritabilities, genetic covariances and number of causal genetic variants while accounting for (and quantifying) uncertainty in genotype-phenotype associations | GEMMA12 |
| Climatic variation | Bayesian modeling of uncertainty in ensembles of climate models | Generates future, predictive distributions of climatic variation with uncertainty over different climate models | JAGS/STAN21 |
| Ecological interactions | N-level structural equation modeling (e.g., generalized linear latent and mixed models (GLLAMM)) | Multilevel extension of structural equation modeling that allows for interactions across hierarchical levels in a Bayesian context; can consider joint uncertainty of model parameters and latent variables | xxM22 |
| Evolution | Forward genetic simulation models (e.g., Wright-Fisher and extensions with age structured populations, etc.) | Flexible models that allow for drift, selection, gene flow, and other evolutionary processes; can be fit in various ways, and can incorporate ecological data | SLiM323 |
| Time series | Autoregressive moving average models (ARMA) | Models that account for spatial or temporal autocorrelation; of broad and general use for time-series analysis | JAGS/STAN24 |
| Combination of data types | Hierarchical (multilevel) Bayesian models | General class of flexible Bayesian models that can combine disparate types of data to make joint inference of evolutionary processes, considering uncertainty from each source and integrated over sources | JAGS/STAN25 |
We focus mostly on hierarchical (i.e., multilevel) models that can be fit in a Bayesian context. Each model accounts for uncertainty (due to data limits or randomness) in a factor relevant for predicting evolution, but an ideal analysis would combine these components to propagate information and uncertainty across these disparate components. We stress that the examples below are representative, but by no means exhaustive.