Estimating time to the most recent common ancestor (TMRCA): comparison and application of eight methods

Investigating how an ancestral population diverge to give rise to distinct sub-populations remains a fundamental pursuit in population genetics. There is broad consensus for the .Out-of-Africa. hypothesis which states that modern human arose about 200,000 years ago in Africa and spread throughout the continent around 100,000 years ago. This was followed by several waves of major population dispersals across the globe, although the exact nature of the population divergence remains debatable. Existing methods to estimate population divergence time differ in their methodological frameworks and demographic assumptions, and require different types of genetic data as input. These fundamental differences often result in the methods producing inconsistent estimates of the population divergence time, further confounding attempts to robustly uncover the history of human migration, especially when most population genetic studies do not employ multiple methods to estimate the time to the most recent common ancestor (TMRCA). Here we simulated a series of different evolutionary scenarios to compare eight methods (Hayes. method, McEvoy.s method, DADI, MIMAR, GPho-CS, CoalHMM, PSMC and MSMC) for their robustness and accuracy in estimating TMRCA. We subsequently apply all eight methods to estimate the population divergence time between Southeast Asian Malays and South Asian Indians using deep whole-genome sequencing data.


Supplementary Materials Download


  • The supplementary document gives the command line we used for this study and can be downloaded here: Supplementary Document – TMRCA.docx
  • The simulated data we used can be downloaded here: simulated_data.tar
  • The perl scripts we used to apply Hayes. method and McEvoy.s method, as well as python scripts we used to run DADI can be downloaded here:




  • Jin Zhou, Yik-Ying Teo
  • Estimating time to the most recent common ancestor (TMRCA): comparison and application of eight methods



If you have any questions regarding to this study, please send an e-mail to both of the following people:

  • Zhou Jin ( )
  • A/Prof Yik Ying Teo ( )