To follow the discussion below, it would be helpful for reference to open the y-Results page of this web site in a new tab.
In this discussion when I refer to an outcome for one of the 37 marker sites tested in our project, I am referring to the number of repeats of short DNA segments that were detected at that site. The marker sites tested in our study are termed STRs =Short Tandem Repeats. As an example, it can be seen in the y-Results page that for the first marker site (#393) all seven Deweeses have 14 copies of the short DNA segment at this site. See below for Technical details of site #393 on the human Y chromosome.
The outcomes for each of the 37 marker sites are shown for the seven participants in the y-Results page. I have colored coded the outcomes at marker sites for which there were differences among the seven participants. There were seven such marker sites where differences occurred.
Marker sites in the top row colored red are those known to have high mutation rates. In fact, six of the seven marker sites for which there were differences among the seven participants are in this high mutation category. Mutations in these markers produce differences among surname lineages in a genealogical time frame and help explain the Dewees(e) data.
Deduced Ancestral Haplotype (DAH) based on outcomes from Y STR markers
Before discussing results of the Dewees Surname Project, I will offer an explanation of how to estimate the Y chromosome marker site values for the Project’s common male ancestor, in this case Garrett Hendricks de Wees. For STR marker sites on the Y chromosome, an individual’s haplotype is the set of outcomes for the several marker sites tested. Due to chance mutations along the Y chromosome, the haplotypes of descendants of a common ancestor will change over time, deviating from the ancestor’s haplotype. Given the haplotypes of a number of male descendants from a common male ancestor, it is sometimes possible to deduce the haplotype of this distant common ancestor. This is termed the Deduced Ancestral Haplotype (DAH). This is accomplished through the use of genealogical records and the number of mutational events required to work back to the ancestor’s haplotype. This information is useful in interpreting the results of surname projects that utilize STR markers on the Y chromosome.
Analysis of the Dewees(e) Surname Project Results
The Dewees(e) Surname Project assumes that WILLIAM, CORNELIUS, and LEWIS are sons of Garrett H. de Wees. Historical records prove this relationship for William and Cornelius, and strong circumstantial evidence exists for Lewis as the son of Garrett as well. The possibility that Lewis is the son of Garrett is addressed below. In addition, it is assumed that true Dewees paternity exists for all seven Deweeses participating in this surname project. The Y STR data discussed below supports this assumption.
Genealogical records indicate that CD and brothers, AAD and JBD, are fourth cousins once removed who trace their ancestry seven and eight generations, respectively, back to WILLIAM. MFD and JTD are fifth cousins who trace their ancestry eight generations back to LEWIS. ACD and HWD are sixth cousins who are descendants of CORNELIUS, eight generations back. The three lineages shown below are, therefore, assumed to be independent lineages descended from the same common ancestor, Garrett H. de Wees, through his three sons. Genealogical pedigrees for all seven project participants can be seen at the Patriarchs page on this web site.
As shown on the y-Results page, of 37 marker sites tested by FT DNA in the Dewees(e) Surname Project, variation among participants occurred at seven sites shown in the green shaded areas. The values for these seven marker sites are presented in the table below for each participant. The string of numbers for each participant is his haplotype. The participants are arranged into their lineages from the three sons of Garrett H. de Wees. As is expected for brothers, AAD and JBD share identical haplotypes which they inherited from their father.
Marker sites: 458 464a 460 576 570 CDYa CDYb
C. D. 15 11 11 17 19 35 41
J. B. D. 16 11 11 16 19 35 41
A. A. D. 16 11 11 16 19 35 41
M. F. D. 15 11 10 16 20 34 42
J. T. D. 15 12 10 16 20 34 40
A. C. D. 15 11 10 16 20 33 42
H. W. D. 15 11 11 16 20 34 41
DAH 15* 11* 10/11* * 16* 20* 34*** ****
* Among these four marker sites, two of three lineages produced the same outcome which is most likely that of their common ancestor, Garrett.
** For markers 460 and CDYb the two outcomes for the DAH are equally likely events, requiring two mutations at a minimum.
*** 34 is the most likely ancestral value for marker CDYa. This requires a mutation from 34 to 35 in WILLIAM’s lineage and a mutation from 34 to 33 in the ACD lineage.
**** Ancestral value cannot be determined for CDYb, which is reported to have a very high mutation rate http://www.jogg.info/22/Chandler.pdf (John F. Chandler Journal of Genetic Genealogy 2:27-33, 2006)
The following conclusions are based on the possibility of chance mutations among 37 tested marker sites occurring during the eight to nine generations that separate the seven Project participants from their purported common ancestor, Garrett H. de Wees. Variations were detected in only seven marker sites. Due to this small amount of variation and the small sample of Dewees participants, the following conclusions are subject to the effects of chance sampling. With the use of more marker sites and additional participants, conclusions could vary somewhat from those presented below.
1. In WILLIAM’S lineage, CD and AAD/JBD differ in two of 37 tested marker sites, #458 and #576.
This level of similarity in Y chromosome makeup is within the range predicted by genetic theory for fourth cousins. Based on theoretical considerations published by Bruce Walsh, the Time back to a Most Recent Common Ancestor (TMRCA) between two relatives, given a difference of 2 out of 37 tested marker sites, can be estimated for specified mutation rates. Use of the TMRCA Calculator at http://www.dnacalculator.org/tmrcaCalculator.php permits the following probability statement to be made for the estimated TMRCA: “there is a 47% chance that the two individuals (such as CD and AAD/JBD) shared a common paternal ancestor within the last 9 generations”. This calculation of “cumulative probability” assumes an average mutation rate of 0.004 for the 37 marker sites. The TMRCA Calculator also reports “…the two individuals most likely shared a common paternal ancestor 7 generations ago”. The actual number of generations back to their common ancestor for CD was five and for brothers AAD and JBD it was six. These observed numbers of generations of five and six compare closely with the theoretical prediction of 7 generations.
2. As is the case for WILLIAM’S descendants in (1.) above, the two descendants of Lewis, MFD and JTD, also differ in only two of the 37 DNA marker sites. Therefore, the calculations used above also apply to MFD and JTD such that two such individuals likely shared a common ancestor 7 generations ago. Being fifth cousins, MFD and JTD shared a common ancestor six generations back, which agrees closely with the theoretical prediction of 7 generations.
3. In CORNELIUS’S lineage, ACD and HWD are sixth cousins who differ in three of 37 tested marker sites, #450, CDYa, and CDYb. Given this level of difference and use of the TMRCA Calculator as above, “there is a 50% chance that two such individuals shared a common paternal ancestor within the last 13 generations”. And, from the TMRCA Calculator,”… the two individuals most likely shared a common paternal ancestor 11 generations ago”. These theoretical values calculated for TMRCA are based on an average mutation rate of 0.004 and are, therefore, over estimates because of the reported high mutation rates of the CDY DNA marker sites. In light of this, the observed differences (at three marker sites) between these sixth cousins agrees with the theoretical prediction within the limits of chance sampling.
An interesting note for these two relatively closely related Deweeses; ACD spells his surname Dewees and HWD spells his DeWeese. As more of Cornelius’ descendants join this surname project, it may be possible to assign unique, deduced haplotypes to these Dewees and DeWeese sub-lineages.
4. In Dewees genealogy, the question of Lewis Dewees’ place in the Dewees family has been an issue. Was he a brother to William and Cornelius as suggested but not proved by genealogical records? To help answer this question, the haplotypes represented in the three Dewees lineages in the above table (WILLIAM, LEWIS, and CORNELIUS) can be compared to the deduced ancestral haplotype (DAH) of their common ancestor, Garrett H de Wees. If the haplotypes of Lewis’s descendants are similar to the DAH, then this would argue that Lewis could be Garrett’s son. Because of the very high mutation rate of CDYb, as referenced above and its ambiguous results in this data set, I have removed it from this discussion.
The haplotype of Lewis descendant, MFD, is most similar to the DAH. He differs in at most one marker site from the DAH. His fifth cousin, JTD, differs in at most two sites. Similarly, in Cornelius’s line, HWD differs from the DAH by at most one marker site and his sixth cousin ,ACD, by as many as two marker sites. Whereas descendants of William, CD and AAD/JBD, differ from the DAH by as many as four marker sites. This information places Lewis’s descendants as close or closer to the DAH (Garrett H. deWees) as descendants of the two verified sons of Garrett H. deWees, Cornelius and William. These comparisons do not prove Lewis to be a son of Garrett H. de Wees, but they do provide strong support to that possibility. The data also would support Lewis as a nephew or other close male relative of Garrett Hendricks deWees. For a thorough review of the historic records that suggest that Lewis is the son of Garrett Hendricks deWees, see the article by Dr. Jack C. Vaughn and Ted D. DeWeese at: http://boards.ancestry.com/surnames.deweese/358/mb.ashx
5. Dewees family historians report that virtually all U. S. Dewees(e) families studied in detail can be traced back to the Garrett Hendricks deWees family who immigrated from The Netherlands to New Amsterdam (now NYC) in approximately 1663. As more Deweeses join the Dewees Surname Project, thus providing additional Y STR data sets, it may soon be possible to assign a Dewees(e) male to his correct lineage and sub-lineage based solely on his Y chromosome makeup.
6. The Dewees(e) Haplogroup
As noted in the Y-Results page, three Dewees males are confirmed by SNP (Single Nucleotide Polymorphism) testing to belong to the I2b1haplogroup. Their I2b1 designation is colored green. The I2b1 haplogroup designations for the remaining four Deweeses are in red and “have been predicted by Family Tree DNA based on unambiguous results”, primarily from their STR haplotypes. The I2b1 haplogroup identifies a group of people who all share the M223 SNP, a rare mutation that occurred in their common male ancestor several thousand years ago. Males having a Y chromosome with this mutation were present in northern Europe after the last ice age and today are concentrated in the countries of Denmark, Germany and the Netherlands. The common ancestor of U. S. Dewees(e) lineages is Garrett Hendricks deWees, who immigrated from the Netherlands to New Amsterdam (now NYC) in about 1663. This information is in agreement with the Dewees(e) haplogroup designation. Learn more about Y haplogroups at http://www.genebase.com/tutorial/item.php?tuId=12.
Andrew A Dewees
Administrator, Dewees(e) Surname Project
Revised March 25, 2012
Technical details of site #393 on the human Y chromosome:
As noted above, site #393 on the Y chromosome (DYS393) of Dewees males in this project contains 14 copies of a short repeating DNA segment. Through the use of nucleotide sequencing techniques, geneticists have determined that the short repeating sequence at this site is AGAT (http://www.cstl.nist.gov/strbase/str_y393.htm). Therefore, under the nucleotide pairing rule (A-T and G-C), the double stranded DNA at this location in these Dewees males appears as follows:
1 2 3 4 5 6 7 8 9 10 11 12 13 14
5’ XXXXXXXXXXXXXX AGAT AGAT AGAT AGAT AGAT AGAT AGAT AGAT AGAT AGAT AGAT AGAT AGAT AGAT XXXXXXXX 3’
3’ XXXXXXXXXXXXXX TCTA TCTA TCTA TCTA TCTA TCTA TCTA TCTA TCTA TCTA TCTA TCTA TCTA TCTA XXXXXXX 5’
On either end of this repeating sequence are nucleotide sequences (xxxxxx…..xxxxx) unique to the DYS393 region that are used to identify and amplify this entire sequence from a cheek cell DNA sample. The resulting pool of amplified DNA molecules from this site is then separated from other DNA molecules and sized to determine the number of repeat copies present at this STR site.
Among males sampled world-wide, the number of AGAT repeats at DYS393 is reported to vary from 9 to 17. This degree of variability is not unusual for STR DNA marker sites and is the basis for their use in characterizing surname lineages in genealogy studies. The variation found in surname lineages for STR markers such as DYS393 is most likely due to occasional errors in DNA replication, which usually increase or decrease the copy number by one. As evidence of this, all seven mutations in this study resulted in the increase or decrease of one repeat unit from the modal number of copies (see y-Results page).
Andrew A Dewees