A Statistical Framework to Infer the Mutation Model of Tandem Repeat Variants.

Area of Science:

• Population Genetics
• Bioinformatics
• Genomics

Background:

• Tandem repeats (TRs) exhibit complex mutation patterns influenced by locus-specific properties.
• Understanding TR mutation models is crucial for deciphering genetic diversity patterns.
• Accurate characterization of TR evolution requires robust mutation models.

Purpose of the Study:

• To develop a computational method, TRAMA, for estimating TR mutation processes.
• To leverage ancestral recombination graph (ARG) information for mutation parameter estimation.
• To compare the Stepwise Mutation Model (SMM) and Two-Phase Mutation Model (TPM) for TR evolution.

Main Methods:

• TRAMA utilizes genealogical history from ARGs to estimate TR mutation parameters.
• The method estimates parameters for both SMM and TPM.
• Model selection is performed to identify the best-fitting model (SMM vs. TPM).

Main Results:

• TRAMA provides accurate mutation rate estimates for TRs under SMM, especially for rates > 10^-5.
• Reasonable estimates for TPM parameters are achieved under specific conditions.
• TRAMA accurately distinguishes between SMM and TPM as the better explanatory model for TR genetic diversity.

Conclusions:

• TRAMA is an effective tool for characterizing TR mutation models using ARG data.
• The method demonstrates accuracy in mutation rate estimation and model selection.
• Estimated mutation rates using inferred ARGs (via SINGER) are comparable to using true histories.