Version 3, August 8, 2011

Step 1) Select one of the larger groupings for analysis

DNA analysis requires large sample sizes of at least five to ten submissions. This grouping of submissions with R1b1a2 estimated haplogroups is the second largest grouping in the Casey DNA Project. None of these submissions in this grouping appear to be closely related in a genealogical time frame (pretty diverse DNA) but many could be related in the distant past (around 600 years ago when Caseys first started using surnames).

Step 2) Determine grouping of submissions

There are twelve submissions in this grouping that have 25 or more common markers to analyze. Due to the diversity of the DNA values, this grouping of submissions should not be considered genealogically related as their DNA varies too significantly. These submissions are analyzed together simply because these all appear to have similar deep ancestry (R1b1a2) and many submissions appear to have a reasonable chance of having a common male ancestor when our ancestors first started using surnames (around 600 years ago). For the seven submissions with 37 or more markers tested, the mutational difference of these submissions varies from nine mutations to fourteen mutations. At 12 generations (time frame where genealogical connections can be made), the highest probability of being related to the estimated MRCA haplotype is a mere 1.18 %. This confirms that none of these submissions in this grouping will be related in a genealogical time frame.

Below is a summary of the twelve "25 or higher" marker submissions as shown by the FTDNA MRCA utility. No submissions matched the MRCA haplotype and the closest matching submission was submission 56479. Submission 56479 will be used to approximate the MRCA haplotype (since FTDNA does not allow manual entry of markers into their database). The FTDNA MRCA utility only allows comparisons of real submissions. At 12 generations, the genetic variation is too significant for any submissions to be closely related:

There are also three additional submissions at 12 markers that could be part of this grouping. With only 12 markers tested, it is not even possible to reliably assign these submissions to any grouping as there is insufficient information to verify their grouping with any certainty:

Check to see if other clusters are related to this cluster (at 12 generations). At 25, 37 and 67 markers, the degree of relatedness is less than one percent. This table below clearly shows that no genealogical connections are possible with other two major clusters (all probabilities are less than one percent):

The submission of the Christopher Casey line (56479) is the closest match to the MRCA of this grouping. This submission is five mutations from the true MRCA haplotype at both 37 markers and 67 markers. Since FTDNA does not allow the manual entry of the true MRCA haplotype, submission 56479 has to be used as the best approximation of the MRCA haplotype. At 37 markers, other submissions range from 7 to 12 mutations from the MRCA and at 67 markers, other submissions range from 12 to 16 mutations.

Since most of these submissions are very distantly related (600 to 1,200 year time frame), there is a pretty high possiblity that a few submissions could have different deep ancestries. If several submissions test for deep ancestry and a new deep ancestry is revealed, this would split this grouping into multiple groupings and create multiple MRCAs for each grouping and would allow more accurate analysis.

Step 3) Look for NPE submissions to be added to cluster

Searching the Y-Search database for matches using submission 56479 (Y-Search 656SG) and using 32 markers and six mutations, 72 matches were found. 32 markers is the common subset between the 37 marker FTDNA test and the Ancestry.com submissions. Six mutations is the maximum mutational difference allowed in Y-Search. There were 72 matching submissions spread across 60 surnames. This wide variation in surnames suggests very common DNA values. Most of these surnames were also common surnames. There were five surnames with more than one submission: Lee (5), Shelton (4), Vance (3), Hamilton (2) and Ross (2). This grouping of submissions must be dominated by "overlapping haplotypes" (due to common DNA values) vs. possible NPE connections.

Unique values for this grouping include (uniqueness of DNA fingerprint):

Having a Delta Weight of only 10 is extremely low indicating extremely common DNA values and imply a very high probability of overlapping haplotypes. The total weight of 378 is extremely low as well. The MRCA of this grouping has only a two mutations from the most common R1b marker values: 449 (29 to 30) and CDYa (37 to 36). For 449 = 29, the percentage is 41 % (only nine higher weighting than this MRCA). For CDYa = 36, the percentage is 30 % (only one higher weighting than this MRCA).

The R1b haplogroup (M343) is believed to have originated around 30,000 years ago yet the MRCA of R1b is only two mutations from the MRCA of this grouping (56479). At 67 markers, the FTDNA MRCA utility indicates that two submissions with a mutational difference of two will be 97 % related at 300 years and 99.9 % related at 600 years. So how can there be such a major disconnect in the MRCA calculation from the well established date of the the M343 (R1b) mutation? The FTDNA MRCA calculator states that 56479 has 99.9 % of being related at 600 years when the M343 haplotype is estimated to occured 30,000 years ago. It is due to common DNA values (overlapping haplotypes). MRCA calculators are simply not very accurate for common DNA values. The MRCA calculators are only accurate when the DNA fingerprint of the MRCA haplotype is not very common (has several rare marker values) or when genealogical evidence (having the same surname that is not extremely common) compensates for having common DNA marker values.

DNA changes over time are similar to the flow of water in a great river. The bulk of the water primarily flows down the deeper parts of the river but some water flows in the shallower portions of the river. The water flows back and forth between the deeper and shallower parts of the river but most water returns to the center part of the river over time. Each marker is like a river and having rarer marker values is like being out in the shallow portion of the river for a period time. Like the flowing water of a river, marker values tend to return to the most common values (in the middle of the river).

Step 4) Looking at submissions at 600 years

This step evaluates to see if this grouping can share a common Casey ancestor. Even at 24 generations, the highest degree of related to submission 56479 (best representative of the MRCA) is only 38 %. Since the selection for the MRCA haplotype is four mutations from the estimated MRCA, these percentages could increase substantially if a new submission becomes available that is more closely related to the true MRCA haplotype:

By upgrading to 67 markers (or even 111 markers), more resolution of testing will show one of three scenarios: 1) improved relatedness; 2) relatedness remains about the same; 3) decreased relatedness. There are only three submissions that tested at 67 markers. Unfortunately, comparison of this resolution (67 markers) the the previous resolutoin (37 markers) shows that degree of relatedness did not change dramatically by upgrading to 67 markers and actually became little less related. However, higher resolution testing will always reveals more accurate conclusions and sample size of only three submissions is too small to conduct accurate analysis.

With such inconclusive information about this grouping of submissions, it is difficult to advise on what the next steps should be. However, there is some additional testing that could help reveal real genealogical clusters that are currently found in this grouping. Deep ancestral tests reveal mutations that happened 2,000 to 20,000 years ago. For a grouping of submissions in this analysis, testing deep ancestry could prove very useful. If you do not share deep ancestral roots, it is virtually impossible to share a common ancestor in the time frame when our Irish ancestors started using surnames around 600 years ago. There are two options for testing deep ancestry. Submission 56479 has tested Y-SNP DNA (deep clade) and resulted in the haplogroup R1b1a2a1a1b5a (positive for SRY2627).

The 37 marker or higher submissions should order a deep clade test. There are two options which I feel obligated to suggest. If you are a gambler and believe that this grouping just beat the odds and mutated more than randomly, you can test just one marker (SRY2627) for only $29 (under advanced SNP tests). If this marker tests positive, you know that you have the same deep ancestry that submission 56479 has. If this test comes back negative, you did not learn anything about your deep ancestry. If you are more conservative, you can order the full deep clade test for $89. Even after you order the deep clade test, you may still have to order individual SNP markers to determine the latest haplogroup (FTDNA lags the latest additions of haplogroups as new haplogroups are being found every month).

Testing deep ancestry of submissions in this grouping could reveal two or three haplogroups which would reveal several genealogical clusters in this grouping. The deep clade tests may allow us to break up this grouping into several groupings. Again, if you do not share the same deep ancestry, you can not share a common ancestor after the time frame where our ancestors started using surnames. The fewer the number of mutations between your submission and submission 56479, the more likely that testing only SRY2627 would test positive. Sponsoring DNA testing of new possibly related lines still remains the best payback at this stage of analysis.

The best usage of funds for this grouping is to increase the number of submissions that have 37 or more markers. This allows for more accurate analysis. This can be done by upgrading some of the submissions that have already tested for fewer than 37 markers. Another source of 37 marker submissions will be new submissions that get assigned to this grouping. You can either wait for random submissions to happen (slow but steady) or you can be proactive and recruit submissions of Casey lines that you suspect are related to your line. Eventually, some new submissions will match existing submissions.

A another usage of funds would be to upgrade submissions to higher resolutions of 67 markers and 111 markers. The more markers tested, the more accurate the analysis will be. Since submissions in this grouping will at most only share common ancestors 400 to 600 years ago, many more markers will be required to establish connections between these submissions in this more distant time frame.

Step 5) - Determine MRCA Haplotype

The most common methodology for determining the MRCA haplotype is the common used "majority rules" methodology. It is assumed that the MRCA would contain DNA marker values that the majority of submissions in the cluster have.

This cluster is very easy to determine the most likely haplotype of the MRCA using the "majority rules" methodology with the exception of two or three markers where the MRCA marker value is not so clear. The marker values for this cluster are shown in the DNA Results summary chart:

DNA Results (R1b1a2 grouping)

Note 1) Since this grouping can not all be related even at 24 generations, the mutations found in this grouping can not reveal possible branches. Due to the low degree of relatedness, estimating any branches would be very problematic.

Note 2) It is unfortunate that FTDNA markers and Genealogy.com markers are not more closely aligned and that economical upgrades are not available for missing markers. It is unfortunate that each company does not at least offer economical upgrades to each other offerings. With the recent acquisition of DNA Heritage by Family Tree DNA, upgrading to common sets of markers may be offered in the near future.

Step 6) - Analyzing Branches within the cluster

Since this grouping of submissions are not closely related enough to be considered a true genealogical cluster, researching common mutations for genetic branches can not be done since this grouping includes many unrelated lines. Since these submissions are not related, no branches can be found at this stage of the analysis (unless they are branches that decend from the same well proven ancestor).

Step 7)

Conclusions that can be reached with current submissions:

1) DNA evidence does not support that any of the submissions in this grouping are closely related in a genealogically significant time frame (200 to 300 years). These submissions do not have a reasonable chance of later being connected together by a combination of DNA information and traditional research. These submissions should not share traditional genealogical research in an attempt to reveal how these submissions are connected.

2) This cluster is not a very unique grouping from a DNA point of view. This grouping does contain eleven lines with the Casey surname. Using submission 56479 as the closest match to the MRCA haplotype, Y-Search reveals 72 submissions within six mutations (32 or more marker values). This grouping shows major evidence of overlapping haplotypes (DNA submissions with common marker values that overlap with other non-Casey surname clusters).

3) Analysis of DNA submissions require traditional documentation associated with each submission. This grouping of submissions does not have traditional documentaion on many submisisons. Please review the current family trees associated with the ancestries of each submission. Below are the known family histories for this grouping (please submit more documentation for the submissions with little documentation available):

Family Trees (R1b1a2 grouping)

Possible Action Items for grouping:

1) Researchers of this Casey grouping should seek out currently non-related Casey lines that they feel could be connected to their line and get their DNA submissions. Around half of these submissions will mostly prove unrelated - but this will eliminate possible related lines that are no longer worthy of attempting to connect to this cluster.

2) This grouping of submissions would benefit from deep ancestry testing. This will help break up this grouping into multiple groupings. Either order only marker SRY2627 at $29 (but risk that no deep ancestry may be revealed or order the full deep clade test at $89 if you are not a risk taker).

3) This grouping definitely needs more traditional information concerning the Casey submissions in this grouping. Please look Casey Family Histories and send the admin more detail concerning your lines. Without more complete traditional information as outlined in the introduction of the Family History (How to Submit) section, a complete analysis of this cluster will not be possible.

4) Since there are many mutations between the MRCA haplotype of this grouping and most of the submissions in this grouping, it would be beneficial to obtain a second son of the oldest proven ancestor of each submission. These additional submissions would assist in the determination of whether the existing mutations are from the older generations or more recent mutations nearer the donor (and are not genealogically significant mutations).

5) Submissions with lower than 37 markers should be upgraded to 37 or more markers. Submissions tested at 37 or more markers could also benefit from upgrading to either 67 or 111 markers. Since most upgrades to 67 markers did not change dramatically the analysis when compared to 37 markers, these upgrades may not reveal much new information. However, there are only three 67 marker submissions and no 111 marker which is a very small sample size which reduces the accuracy of any conclusions..