How Many Markers Should I Test

The question asked the most often by people considering obtaining the Y-DNA test is: “How many markers are enough?” “After all, the Y-DNA test is available for 12, 25, 37 and 67 markers and the more markers tested, the higher the cost of the test. Which test is most cost effective for my situation?”

The place to start to answer this question is to understand the term ‘marker'. A marker is a definitive location on the Y-chromosome that may be tested for genetic genealogical purposes. These locations, or markers, have names, such as DYS #19 or DYS #385a or DYS #439. When a marker is tested, the result is reduced to a number, which represents the number of repeated patterns (known as alleles) of the DNA protein sequence at that specific location on the Y-chromosome.

Family Tree DNA offers a 12-marker Y-DNA test, a 25-marker Y-DNA test, a 37-marker Y-DNA test, a 67-marker test and now some additional markers above 67 on an ala carte basis. A test with a higher number of markers will produce results, that reduces the time frame to the Most Recent Common Ancestor, or MRCA. For all four of the standard panels of tests (12, 25, 37 & 67), the number of markers that match can determine whether you and another participant share a common ancestor and how many generations ago that common ancestor might have lived.

Basically, you want to select the test that addresses your immediate situation; perhaps that’s the least expensive to get started. It is important to understand that once you have had a sample of your (or your relative’s) DNA tested by FTDNA, you can obtain a more definitive test (upgrade to more markers) or even another type of test (such as an mtDNA test) later on (up to 25 years later without a new swab required). However, if you order a Y-DNA test for the higher number of markers initially, you save a few dollars as compared to initially ordering a 12-marker test and later upgrading to a higher number of markers at a later date and you don't have to wait a second time for your results to be posted. See the FTDNA Test Price Table.

For the Knowles Surname Project, there are three major objectives;

1. For the descendants of all Knowles Progenitors, determine and prove who is related to whom,

2. Determine if any of the known Knowles Progenitors have a common ancestor,

3. Determine which Knowles Progenitor is the ancestor for those Knowles families who are unaware of which Knowles line they belong to.

Surname Projects vary in their size and scope, as does the objectives and will depend on the frequency of the surname. Some Surname Projects focus on specific lines, immigrants, families, or countries, while the objectives of other Surname Projects focus on validating research. Often the frequency of the surname has an impact on the determination of the project objectives. High frequency surnames may be run in phases or groups to manage the number of participants. Low frequency, or rare surnames, may be able to address the total population in one phase. The Knowles Surname Project is somewhere in between these extremes; i.e., Knowles is not extremely rare nor is the challenge of a surname like Smith, Johnson, Jones, etc.

Because of the Knowles Surname Project objectives, ideally participants that can afford it are requested to obtain at least a 25-marker test, with the 37-marker test preferred. However, for those participants that have some doubt as to their Knowles line, initially obtaining the results of a 12-marker test is certainly warranted.

Under any testing scenario, the results for a participant may suggest more genealogical research, or point to a specific area, country or hopefully to a specific ancestor.

For a Regional DNA Project, such as the Wiregrass Georgia or the Lumbee Tribe projects (also managed by Robert B. Noles), you are trying to determine if your Y-DNA profile matches others in the same project. For example, if your surname is SMITH and your Smith ancestors are known to be from the Wiregrass Georgia area for many years, you might be better off participating in the Wiregrass Georgia Regional DNA Project than a Smith Surname Project that has Smith participants from all of the U.S. Another reason to belong to a Regional DNA Project is when you suspect there has been an undocumented surname change or unknown surname ancestor in previous generations. The more uncertain you are about your genealogy and in those cases where a surname change has occurred, the more markers tested the better.

The following paragraphs are an attempt to help you understand how the results of tests conducted at the different marker levels compare.

If two individual's test results match exactly (12/12) in the 12-marker test, there is a 99% probability that they are related. The issue then becomes: when did this common ancestor live? Unfortunately, science cannot yet pinpoint the exact generation, but science can provide a range of time when the common ancestor might have lived. This is when you need to use your knowledge of your ancestor’s family tree as determined via traditional research.

If two individuals match in the 12-marker test for either 10 out of 12 (10/12) or 11 out of 12 (11/12), they are also considered related, but the time frame to the common ancestor, MRCA, is more distant than if they had a 12/12 match. Where the matches are less that 10/12, the two individuals are not considered to be related; certainly not in genealogical terms.

If your 12-marker test results match another participant’s exactly, 12/12, your common ancestor occurred between 1 and 62 generations ago, with a 50% probability that the common ancestor lived 14.5 generations ago or less. There is a 90% probability it was within 48 generations and a 95% probability it was within 62.

You can shorten this time span to the MRCA as determined by the 12-marker test significantly by increasing the number of markers tested to 25 or 37.

If two individuals match exactly (25/25) in the 25-marker test, their MRCA would have lived between 1 and 32 generations ago, with a 50% probability that the common ancestor lived 7 generations ago or less. There is a 90% probability the MRCA was within 24 generations and a 95% probability that it was within 32 generations. Therefore, increasing the markers tested from 12 to 25 lowers the time frame to the most likely MRCA from 14.5 to 7 generations.

If two individuals match exactly (37/37) in the 37-marker test, their MRCA would have lived between 1 and 20 generations ago, with a 50% probability that the common ancestor lived 5 generations ago or less. There is a 90% probability the MRCA was within 16 generations and a 95% probability that it was within 20 generations. Therefore, increasing the markers tested from 12 to 37 lowers the most likely time frame to the MRCA from 14.5 to 5 generations.

NOTE: Graphs showing the Time to the Most Recent Common Ancestor for the 12, 25 and 37 marker tests, and the associated confidence levels are within the TMRCA article included in under DNA - 101. The TMRCA also discusses the probabilities and number of generations to the MRCA for situations where the two individuals have less than a perfect match.

In the future, there may be tests available that process even more markers; perhaps as many as 100, before we achieve diminishing returns. The testing of even more markers will no doubt be made available when more genealogists are looking for even greater accuracy and proof that two people descend from a recent common ancestor. If two individuals match exactly (100/100) in a 100-marker test, the time frame to the MRCA collapses from the 14.5 generations for a 12-marker test to 1.7 generations for the 100-marker test.

Participants in a Surname Project choose the number of markers to be tested depending on their objectives and the objective of the Project. A 12-marker test can be upgraded to a 25-marker or a 37-marker test at a later date. The testing lab used by Family Tree DNA, based at the University of Arizona in Tucson, keeps a sample of your DNA stored under a kit number at an off-site secure location. If a participant in a DNA Project initially chooses a 12-marker test and then later decides to expand their test to 25 or 37 markers, the additional markers can easily be tested from the DNA already stored for reprocessing. No new swab is required.

Surname Projects participants often struggle with the choice between 12, 25 and 37 markers, as well as who in the family should be tested (assuming there is more than one candidate available). The recommendations for the Knowles Surname Project participants varies based on the participants objectives, the situation as to available candidates and the potential benefit to the overall project.

Examples of different scenarios are discussed below to assist you in making a decision for the number of markers and selecting a participant from your family, whether your participation is for the Knowles Surname Project or you are considering a project for another family surname.

Most likely the test results for some participants will show they are not related. By starting off with the 12-marker test, those participants who are not related do not purchase any testing that is not needed. The 12-marker test will clearly identify those persons who are definitely not related. For those participants whose test identifies that they are related, the next step is typically an upgrade to the 25 or 37-marker test. As discussed above the upgrade to 25, 37 or 67 markers will reduce the time frame of the common ancestor, and identify the degree of relatedness of the participants. Here are a few example situations:

Situation 1 - There is only one surviving direct male descendent in your Knowles line, traced back to the early 1800's with documented research, but no further. If the surname of concern is not Knowles, you are not ready to start a surname project and haven't researched other lines with your surname. You never find the time to spend to understand DNA testing. Your direct male descendent is in his late 40's.

Recommendation: 12 Marker Test - On the surface, it appears that you have plenty of time to learn about DNA testing, and have your relative tested. Unfortunately, once this direct male descendent is gone -- you can probably never get a DNA sample. In the situation where there is only one surviving direct male descendent, regardless of their age, immediate testing is recommended. If financial constraints are a problem, the 12-marker test will do. Any time in the future, if additional direct male descendants are uncovered from your Knowles line (or from the other surname) they can be approached to be tested for comparison. The objective today is to get a sample from your relative and have it tested to ensure that you will not be denied the opportunity in the future. You will then have the results for the direct male descendent, the results will be stored in the FTDNA database, and the sample will be stored for your future use.

If there are as few as three direct male descendents in your tree, immediate testing of two of the males is highly recommended. The reason to select two is to confirm that there were no adoptions or extramarital male births in one or both of the lines.

Often, it is only a priority to test males if they are elderly. However, the issue is not age. If the person is gone, for all practical purposes, their Y-DNA is gone. Family Tree DNA was able to assist a client, whose only direct male descendent was killed in a car accident. However, securing the sample took tremendous effort, and the involvement of the spouse for approval and the medical examiner for a sample. It was sheer luck that the client remembered in such a stressful situation, and contacted FTDNA on a Saturday night at midnight, so that they could immediately overnight a collection kit. You don’t really want to be in this type of situation.

In summary, if you have a small number of direct male descendents in your line, we recommend immediate testing of at least two participants, to confirm a match and to have the sample and results stored for future reference. If you do not get an exact match with the initial participants, expand the testing to any remaining direct male descendents to determine where and when a non-paternity event occurred, such as adoption.

Situation 2 - Your research has established that two immigrants arrived in the U.S. in 1740 with the same surname; perhaps they were brothers or cousins. Extensive research has occurred, but you can't connect these two immigrants with traditional genealogical paper records. There are many descendents today in the U.S. from both of these immigrants. This situation may apply to a Knowles line or another surname that you are researching.

Recommendation: 12-Marker Test - Testing of at least two of the direct male descendents from each immigrant's line for 12-markers is recommended. The rational behind this recommendation is that there is large span of years between the immigrants’ arrival and today, as well as many birth events that provide an opportunity for an extramarital male birth or an adoption not detected by the genealogical research.

By selecting two males from each line to test, you would expect a match within each line to validate the results. Then, when you compare the results from the two lines to each other, you would be confident that you have meaningful results at a minimum of expense. If all the participants are an exact 12 for 12 match, you may want to increase the number of markers tested to 25 or 37, depending on your need for accuracy and the establishment of the time frame to the common ancestor of the immigrants.

If in this example, the 12-marker test results are 12 for 12 within both lines, but do not match the other line by at least 10 for 12, then you have proven that the two immigrants were not related. If the lines compare at the 10 for 12 or 11 for 12, you will want to upgrade the testing to the 25, 37 or 67 marker level to establish if the two immigrants were related.

If the test for the two descendants within the same line fail to achieve at least a 10 for 12 match, then these individuals are not related genetically. Some mistake or omission in your research has been made. In this case, you will want to find another couple of participants from that line to test at the 12-marker level in order to establish where your research is inaccurate.

Situation 3 - You are interested in determining which lines for your surname are related or not. Your surname can be found in England, Ireland and the U.S. You have extensive paper documentation on your line and have contact with other researchers for your surname in U.S. and in England who have researched their line. You and the other researchers have uncovered literally dozens of apparently different, unrelated ancestors (progenitors) for your surname. Where and how do you start? Note: This is exactly the situation we have for the Knowles surname.

Recommendation: This situation is an excellent candidate for the establishment for a Surname DNA Project (as has been done for the Knowles surname, and the many variants for the Knowles surname, including, but not limited to Knoles, Knolls, Noles, Nowels, etc.).

If a project for your surname of interest exists, you should join that project. If a project for your surname doesn't exist, a Surname Project can be established with a minimum of six participants, or less if approved by FTDNA. If you are someone from your surname group doesn't feel up to managing a surname project, creating a Web site for the project, recruiting participants for the project, etc., you need to look for an appropriate Regional DNA Project that already exists to join.

There are two primary approaches, depending on the potential size of the Surname Project and the extent of the paper genealogical records available. For a relatively rare surname, with limited descendents and excellent genealogical records, testing one or two direct male descendents from each line for 12 markers is initially recommended.

If matches occur, then it may be necessary to expand testing to 25, 37 or 67 markers for the original participants, depending on the need for accuracy and the establishment of the time frame to the common ancestor for the lines that now appear to be related.

When no match occurs between two participants from the same line, an additional direct male descendent should be tested from each of the non-matching lines.

For surnames that are not rare and that have many branches and descendents, testing two males from each line is recommended to establish the genetic fingerprint or profile for each independent line. (Remember: In every generation, the opportunity for an extramarital male birth or adoption exists). Where no match occurs, we will need to expand the testing in the two lines that do not match to include an additional male in each line.

CONCLUSIONS

In conclusion, the selection of how many markers to test in an initial series of tests, depends on the situations and the objectives that will vary between surnames and families. The example scenarios discussed above may help you determine what approach to take with your DNA test situation.

If you are not sure of your objectives, but just want to get started, start by obtaining the 12-marker test, and then later upgrade to the 25, 37 or 67-marker test when a match occurs, to reduce the time frame to the common ancestor, MRCA.

If you are considering starting your own Surname Project, Family Tree DNA representatives are available via email and phone to discuss your situation and to help you determine if a 12-marker, 25-marker or 37-marker test is best for you and your group based on your specific situation and objectives. If you would like an analysis and recommendation for your surname situation from FTDNA, send an email to: bcg@familytreeDNA.com explaining the data known about the genealogy for your surname and the objectives that you have for DNA testing. FTDNA will promptly provide their feedback and suggestions for your situation.

If your situation concerns the Knowles, Marvel, or Prettyman Surname DNA Projects or the Wiregrass Georgia (southeastern Georgia), Roanoke Lost Colony, or the Lumbee Tribe Regional DNA Project or one of the following Wiregrass Georgia Region Surname Projects [Bagley, Courson, Hersey (Hursey), Knowles (Noles), Paulk, Rowell, Spell(s), and Surrency], send your inquiry to Robert B. Noles at: rbnoles@bellsouth.net. A recommendation for the appropriate test for your situation will be provided by return email.