The aim of this site is to provide genetic genealogy hobbyists with the facility to obtain realistic estimates of the time in generations to the most recent common ancestor of two males (their “TMRCA”).  It is meant to be used for two haplotypes of the same haplogroup, and where kinship is suspected within the last 50-odd generations. 

The estimates are not provided by mathematical formulae.

Rather, they are calculated from the cumulative frequencies with which TMRCA’s ranging from 1 to 60 generations are observed in 200 populations of haplotypes simulated using the Monte Carlo method. The simulation process is described in the section named ‘Simulations’ and an example of the calculations is given in the section named ’Calcs’.     

In all, some 7.4 million comparisons are made. What the code on the website does is to search the 200 simulated populations to identify those pairwise comparisons that have the given inputs. It then ranks the TMRCA’s for those found from low to high and reports the cumulative frequencies for up to and including 5, 10, 15, 20, 25 and 30 generations. It reports these frequencies, plus the number of observations found, and the suggestion is that the reported frequencies are realistic estimates of the probabilities of occurrence. It also reports the mean TMRCA for the given inputs, plus the observed 95% frequency intervals.

At this stage, only bi-directional single-step mutations are allowed for, with probabilities of m/2 up and m/2 down, where m is the assumed marker-specific mutation rate. However, the first major change planned for the site is to include two-step mutations and this should happen within perhaps two months.

The TMRCA look-up facility was added on June 3rd in the section called ‘Test’ and the answers it provides are identical to what my home computer produces.  The section called ‘Test’ has since been renamed ‘Live Lookups’. The layout was improved on 20 June and the mean TMRCA plus 95% frequency intervals have been included as part of the output since 4 July.    

Update February 2012:

The site initially considered only 67-marker haplotypes, but lookup facilities for 111-marker haplotypes were added in (I think) November 2011. At the same time, allowance was made for 2-step mutations at a rate of .035 of all mutations, and simulating larger populations (between 2,000 and 3,000 haplotypes) resulted in some 27 million pairwise comparisons compared to the earlier 7.4 million comparisons.

The section ‘Live Lookups’ has been replaced by ’67 Marker Lookups’ and 111 ‘Marker Lookups’.

Sandy Paterson


Comments are closed.