[an error occurred while processing this directive]
TiP and Beyond
This page explores uses of the advanced time to most recent common
ancestor (TMRCA) calculator that Family Tree DNA calls "TiP".
available only on the the private, secured portions of the FTDNA website,
www.familytreedna.com. This means
it is available only to FTDNA customers and project administrators.
- Individual customers may access TiP after logging on to their personal
"My FTDNA" pages with their user IDs and passwords and conducting a search
for matches. If a match is listed and the matchee hasn't kept his
information private, an orange TiP icon will show next to the name; it's a
link to the calculator.
- Project administrators may access YDNA TiP directly from the "Genetic
Reports" tab on the Group Administrator Pages (GAP). Any two
project members may be compared. It is also available from the "Genetic
Some people are unhappy with the TMRCA concept; reality doesn't always match the predictions.
We need to remember that these calculated probabilities are only maximum
likelihood estimates (MLE) and further that their confidence limits are
unknown. It's fair to say that any TMRCA is not a precise number but a
window with an error of unknown size.
For example, a coin toss is expected to have a 50% probability coming
up heads. But in any series of tosses, there may be more heads or tails. Only as
one increases the tosses to a large number does the percentage of
heads (and tails) approach 50%.
What is TiP?
Like other such calculators, TiP compares two sets (a pair) of Y-DNA STR
results to compute a maximum likelihood estimate (MLE) for a cumulative
probability that the most recent common ancestor (MRCA) lived within a
specified number of generations.
- Flexibility -- TiP will calculate probabilities
- Precision -- Other TMRCA calculators use averages of mutation
frequencies for sets of markers. TiP is based (for the most part) on individual marker mutation
- Additional uses -- TiP is useful for determining relative closeness of
relationship in a list of matches. In general, one is more closely related
to this with higher TiP scores.
- Undocumented -- The algorithm is proprietary and undisclosed; it can not
be independently evaluated.
- Does not account for convergence. About 5% of men have false positive
matches across haplogroups. (No ySTR-based TMRCA calculator handles
TiP utilizes a Gamma distribution model for its probability estimates. (For
more on gammas, see this page.) The two parameters (α & β) for the distribution are
related to the number and type (e.g., marker volatility, palindromic) of
haplotype differences. For more mismatches or mismatches on more stable
markers, the parameters used increase to reduce probability at a given
number of generations. .
Other TMRCA calculators also use Gamma distributions but their parameter
handling is less sophisticated.
TiP can be used in more than one way. For these uses, we recommend:
- Use the "per single generation" option to produce TiP reports for every
generation from 1 to 24. (Do not use the "not related within" option;
this adjustment can be made later.)
- Copy and paste the output to a spreadsheet file (like Excel's)
- Transcribe the text output into numbers-only columns for generations
and cumulative probability.
These steps enable calculations for further use of the data.
Assessing match quality
Your matches may be listed in order of "genetic distance" from you; those at the
top are not necessarily your closest relations. For any given number of generations,
your closest relations will be those with the highest cumulative probability
as given by TiP.
More preparatory steps are needed to focus research to find the identity
of the MRCA
- Calculate per-generation probabilities; use another column
for this: Subtract the
cumulative probability of each generation from its more recent neighbor.
For example subtract the probability for generation 5 from 6 to get a
per-generation estimate for 6.
However, these numbers are not particularly reliable or precise. So
another step is needed.
- Sum individual per-generations probabilities into adjacent groups.
For example, sum 3,4 & 5.
One group of generations will stand out as most probable for where in the
family tree the MRCA is to be found. A group of three generations for
example, yields a time window of about 100 years.