If there is just one thing to learn in order to work effectively with your DNA results, it would be to gain an understanding of how centiMorgans relate to relationship predictions.
In the simplest of terms, testing companies use centiMorgans as a measurement of DNA to denote the size of a match segment in autosomal DNA testing. Those with greater shared centMorgans (cMs) and segments are predicted to be closer relatives than those sharing fewer cMs and segments. Ancestry, FTDNA, and My Heritage quote their matches in cMs, whereas 23andMe quotes a percentage of DNA share.
For a more in depth understanding of the definition of centiMorgan, please see ISOGG
Keep in mind that the number of centiMorgans quoted by each testing company is based upon each testing company’s own algorithms, and those differences can be seen by comparing the same two matches at different testing companies. In general, Ancestry’s cMs might register as slightly lower when compared at another testing company due to Ancestry’s Timber algorithm which eliminates some high frequency segments. On the other end of the spectrum, FTDNA might appear with higher shared cMs because they count small segments that the other companies do not include in their totals. By familiarizing yourself with the relationship predictions at each site and how each company counts that information, you’ll begin to build a better picture of potential relationships with your DNA matches.
The good news is that these estimates are easy to explore! There are charts and tools available to estimate potential relationships when two matches share a given amount of DNA.
The Shared cM Project was spearheaded by Blaine Bettinger as a crowd-sourced gathering of data with thousands of known relationships submitted by users over several years. The data that was collected has been worked and reworked to begin to establish good estimates of expected DNA shares within given levels of relationship. To use the Shared cM Project, you'll simply compare the number of cMs shared between two individuals to see where that fits within categories of relationships.
As you’ll note, the greater the number of cMs shared, the closer the relationship, and the fewer possibilities for those relationships. In general, matches that are 3rd cousin and closer are more accurately estimated than those that are further away. However, statistically, even a small percentage of 3rd cousins may share no DNA at all, whereas some more distant cousins may share more than the "expected" amount of DNA.
Another great tool that should be in every genealogist’s toolbox is the interactive Shared cM Tool at DNA Painter. Web developer and genetic genealogist, Jonny Perl, created an interactive version of Blaine’s chart. It’s free, easy to use, and includes probabilities that you can use to evaluate the most likely relationships based on the amount of DNA that two people share.
To get started:
1. Go to DNA Painter and click on the Tools tab.
2. Click on the Shared cM Tool.
3. When the page loads, please note that the image you see on that page includes the original Blaine Bettinger Shared cM Project image with all possibilities shown..
4. Now here is where the magic begins! You have two options here. At the top of the page:
Input the amount of shared centiMorgans
OR
Click on the percent link to open another box to allow you to input data by percentage shared rather than cMs shared.
When you include the shared cMs or percentage, the data will then be sorted to provide you with the various relationship possibilities as well as probabilities of those relationships. It's that fast!
Let’s explore an example between two known half-first cousins, once removed or 1/2-1C/1R (the grandfather of one is the great-grandfather of the other but through different spouses)
As you can see, the data was sorted to show possible relationships and the probabilities chart shows the likelihood of those various relationships. All of the possible relationships that fall outside of the probabilities are grayed out. The known relationship of half-1C/1R is included in the highest probability of 59.07%, but be aware that because DNA is inherited randomly, cousins at the same level of relationship may show different percentages.
Here is an example of the brother of the previous match to the same cousin:
Because these two full siblings share a 100 cM difference in their matches with their known half-1C/1R, it changes not only the potential relationships, but the probability estimates as well. In this second case, the probability of the known relationship of 1/2-1C/1R drops to 40.7%, the second category. This is a good example of needing to be aware of inheritance differences and how it can affect the estimated relationships between two individuals.
Now that you have the probabilities and potential relationships, what do you do next?
Once you’ve determined probable relationships between two people with an unknown relationship, the next step is to work with trees to figure out the most likely relationship. This is where standard genealogy meets genetic genealogy! You will be guided by the shared matches found at your testing company to determine which part of YOUR tree to explore first. And, you will be guided by shared matches they have to known cousins within THEIR tree. As you work with the data, be sure to keep referring back to the Shared cM Project. It will help keep you focused on potential relationships as you work to solve your genealogical mysteries.
Source Citations:
1. Image: The Shared cM Project. [cited 2019 Aug 9] https://isogg.org/wiki/CentiMorgan
2. Images: DNA Painter Shared cM Project - Interactive [cited 2019 Aug 9] https://dnapainter.com/
#DNA #GeneticGenealogy #DNAMatches #Genealogy #UnderstandingDNAResults
Comments