Table 1. Approaches for identifying differentially methylated loci in bisulfite sequencing data sets.
| Statistical method | Directly models counts? | Controls for biological covariates? | Controls for genetic covariance? | Programs that implement the method |
|---|---|---|---|---|
| t-test or Wilcoxon rank-sum test | No | No | No | R and many others |
| Fisher’s exact test | Yes | No | No | R and many others |
| Binomial regression | Yes | Yes | No | R and many others |
| Linear regression | No | Yes | No | R and many others |
| Beta-binomial model | Yes | Some 1 | No | DSS [31], MOABS [32], RadMeth [33] |
| Linear mixed model | No | Yes | Yes | GEMMA [34], EMMA [35], EMMAX [36], FaST-LMM [37] |
| Binomial mixed model | Yes | Yes | Yes | MACAU |
1Only RadMeth; the implementations of the beta-binomial model in MOABS and DSS do not allow the user to control for covariates.