YDNA Similarity Analysis

Return to The Tapestry Homepage		Return to Methods
Contents 1 Related 2 Similarity Analysis 3 One Kit Comparison 4 Multi kit comparison	This is one of a series of articles on Genealogical Methods, prepared in association with The Tapestry. See Index for a list of related articles.

Related

Similarity Analysis

In YDNA Similarity Analysis multiple haplotypes are compared with each other, on the basis of the number of markers each kit has in common with the other kits. For example, if two kits test at 37 markers, and differ from each other on 2 of those markers, their similarity would be expressed as:

∂' =35/37=94.5% similarity.

The above is equivalent to saying "35 marker match out of 37 markers", or simply a "35 out of 37 match"

This can also be expressed as dissimilarity, as:

∂=2/37=5.5% dissimilarity.

This is commonly referred to as a "2 off of 37 marker match", or simply a 2-off match.

Results of similarity analysis are almost always presented in terms of dissimilarity∂. As a simplification in presentation of results the results of similarity analysis are usually presented as ∂ (dissimilarity) but are commonly referred to as simply "similarity".

One Kit Comparison

Results of similarity analysis are commonly displayed either in terms of

One kit compared to all other kits in a set of kits

Each kit compared with every other kit.

In the former case, the primary focus of the analysis is to see how one particular kit compares to every other kit being considered. Here's an example of this.

The original data set used to generate the above included over 130 separate kits, only 32 of which are shown. The principle advantage of the above approach is that the calculations are relatively fast, and the displays of the data easier to create. This type of analysis is best used when focused on a particular kit, and trying to identify the what other kits share a common ancestor with that kit.

Multi kit comparison

The alternative approach is to compare each and every kit in a set with each and every other kit. This approach is better suited for speaking to questions of project grouping where the desired goal is to isolate all kits that share a relatively recent common ancestor. The disadvantage of this approach is that the calculations take more and more time as the number of kits involved increases. In addition, when the number of kits involved gets very large (say 1000 kits) the displays become very combumbersome. Typically a "matrix approach" is used for these displays. Here's an example based on a sample of kits from the Stockton YDNA project.

Currently, the kits shown have been placed in a single group, the Stockton Cheshire G, based on their YDNA similarity. (Note: Empty cells in the matrix display represent kit comparisons that exceed the acceptance criteria for the "run".)