Page 1

Part I: Analysis of Group 2 Genetic Results

(Click Here to go to Part II)

Introduction

Between 1695 and 1725, three men with the surnames Stark and Starke arrived in New Hampshire and Virginia. Their names were Dr. Richard Starke of York County, Virginia, James Stark of Stafford County, Virginia, and Archibald Stark of New Hampshire (father of General John Stark of Revolutionary War fame). The genealogical research has not been able to determine if these three men were related. However, independent research of each of these men has suggested the ancestral home of each could have been in or near Glasgow, Lanark, Scotland. Many of the early Stark family researchers had claimed the Aaron Stark family and these three men were somehow related. Y-DNA comparisons of members of Group 1 — descendants of Aaron Stark — to members of Group 2, have revealed Aaron Stark was not related to these three later arrivals to America. 

The following summarizes from whom the members of Group 2 descend:

Descendants of Dr. Richard Starke [1668-1704]: N21529.

Descendants of Archibald Stark [1693-1758]: 25347, 94630, and 95073. 

Descendants of James Stark [1695-1754]: 74402, 76284, 76964, 84342, 115170, 115705, 136832, and 171929.

Probable Descendants of James Stark: 115170, 136832, 175209.

The usage of surnames did not become common until the 13th and 14th centuries. By the year 1500, surname usage had become common practice. Considering that one generation is about 25 years or more, then the year 1500 or before would be the approximate equivalent of about 20 generations back from the present generations. As a general rule of relatedness, the following will discuss the probability that two individuals shared a common ancestor within 20 generations. A conservative probability of 80% or greater — for the purposes of this discussion — will be considered sufficient to declare that any two individuals compared; having the surname Stark or one of it's derivatives; most likely have a common ancestor that lived approximately within the last 500 years.

This population of individuals have been variously tested over Panel 1 (Markers 1-12), Panel 2 (Markers 13-25), and Panel 3 (Markers 26-37) as will be presented in the Genetic Results Table. One or more participants can have the same haplotype over any combination of these Panels . For the purposes of this discussion, the most common haplotype over a selected group of DYS Markers will be referred to as Haplotype 1. Other lesser occurring haplotypes will be numbered 2 thru N (Haplotype 2, Haplotype 3,..., Haplotype N). All genetic comparisons of lesser occurring haplotypes in this report will be compared to the most common haplotype.

Terminology

Before we begin, here are some explanations of common terms used in this summary.

Page 2

Genetic Results

The following Genetic Results Table presents the allele values available over 37 DYS Markers for each of the members. Cells with this background color report DYS Marker allele values of descendants of Dr. Richard Starke [1664-1704] of York County, Virginia. Cells with this background color report DYS Marker allele values for descendants of Archibald Stark [1693-1758] of Hew Hampshire. Cells with this background color report DYS Marker allele values for descendants of James Stark [1695-1754] of Stafford County, Virginia. The Common Ancestor Column reveals from which from whom each member descends — according to the genealogical research. The "letter" preceding each Kit # will be the "Letter Code ID" for each Member used when discussing genetic comparisons. [Comparison of A-B, for example.]

For the purpose of this report, a Mutation occurs in a DYS Marker column when an allele value differs from the most common allele value in that column. For example, in Panel 1 (Markers 1 thru 12), all of the allele values in the Marker 1 column are 13 — resulting in no observed mutations in this column. However, in the Marker 2 column, the allele value 24 occurs more often than the allele value 23. Therefore, Members B, C, and D — having the allele value 23 in the Marker 2 column — have their Marker 2 allele values highlighted in yellow; indicating each has a Marker 2 mutation relative to the more common allele value of 13. This method of determining mutations results in a minimum number of mutations for the population tested.

According to FTDNA, those DYS Markers labeled in RED have a higher mutation rate than those Markers not labeled in Red.

To review the lineage of any Member, clicking  on the "Kit#" will take you to a more complete lineage report. If the Allele value in the Marker 37 column is presented with a Red colored font, this member has genetic results beyond 37 markers. Clicking on the Kit# will take you to a presentation of those additional DYS markers.

Haplogroup R-M269 is a shorthand notation for Haplogroup R1b1a2. Red colored font for a member's haplogroup indicates a FTDNA prediction which has not been confirmed.

If the Allele value in the Marker 37 column is presented with a Red colored font, this member has genetic results beyond 37 markers. These results are  presented on the 67 Marker Y-DNA Test Results Web page.

Genetic Results Table

Page 3

Group 2: Genealogical Lineage Table

The Group 2 Lineage Table shows the direct male line from the Kit Number to the earliest known ancestor for that kit. Reading from right to left, the earliest known Ancestor of each member of Group 2 is shown in the left columns (Gen 9 & Gen 10). Dr. Richard Starke and John Stark III of Killermont are the earliest known ancestors reported in the Gen 10 column. James Stark of Stafford, County, VA is the earliest known ancestor in the Gen 9 column. The common ancestor of all members of Group 2 is not known with certainty. The Group 2 Genetic Genealogy Report suggests William Stark of Dullutur [1520-1575] may have been the common ancestor of all — but proof of descent from William needs further genealogical research.

You will note that the participants in Group 2 cover three generations (Gen 0, Gen 1, & Gen 2). The youngest are in Gen 0; the Parent Generation of Gen 0 would be Gen 1; and the grandparent Generation would be Gen 2.

The right most column shows the Panels tested for the member in that row and the number of mismatches in each Panel — relevant to Kit #74402 (In Gen 1 of Row E). Click on the Kit # for a more detailed report for each participant and his ancestors. Use your back button to return to this page.

Page 4

Page 5

Analysis of the H37 Haplotypes

Table 4 identifies a number of unique H37 haplotypes when the markers of Panels 1, 2 & 3 are combined.

Referring to the Genetic Results Table, three members of Group 2 (E, F1, & K) had no mutations over all 37 markers and are a perfect match. This repeating Haplotype has been assigned H37-1 as presented in Table 4. When E, F1, and K are compared to each other, there is a 99.99% probability they share a common ancestor who lived within the last 20 generations. 

These comparisons confirm descendants of James Stark of Stafford County, Virginia, Dr. Richard Starke of York County, Virginia, and Archibald Stark of New Hampshire have a common ancestor who most likely lived after the calendar year 1500 and before the birth of Richard in 1668. 

Conclusions

Genealogical research suggests E is a descendant of James Stark. The perfect H37-1 match of E, F1, & K suggests K is most likely also a descendant of James; but his genealogy is not known at this time. The H12-1 haplotype of E is a perfect match to F, F1, H, I, J, & K. Therefore, all of these Members are genetically related and most likely share James as a common ancestor; as suggested by the genealogical research. The H37-1 haplotype comparison of E to the H37-4 haplotype of G and the H37-8 Haplotype of L was required to genetically confirm G & L  shared a common ancestor with E within 20 generations  — therefore, confirming all of the descendants of James Stark in Group 2 are related although the genealogy is not complete for Members J, K, and L.

The H37-1 haplotype comparison of E to H37-3 haplotype of A genetically confirms descendants of James Stark and Dr. Richard Starke share a common ancestor who lived after the calendar year 1500. Therefore, all of the descendants of James Stark are related to descendants of Richard Starke — the genealogical research unable to determine the name or relationship of their common ancestor.

The H37-1 haplotype comparison of the E to the H37-6 haplotype of D genetically confirms descendants of James Stark share a common ancestor with descendants of Archibald Stark who lived after the calendar year 1500. The H12-2 haplotype comparison of descendants of Archibald Stark revealed they were a perfect match to each other (see Table 2, H12-2) — confirming all of the descendants of Archibald are related to each other and most likely share Archibald as a common ancestor. If D shares a common ancestor with the descendants of James Stark as described above, then D also shares a common ancestor with the descendants of Richard Starke who lived after the calendar year 1500. It then follows that if D is related to descendants of James and Richard; therefore, B & C are related to these same descendants.

The H12-1 haplotype could be a genetic profile identifying future participants who descend from James Stark of Stafford County, Virginia. H12-3 (G) may be a genetic profile identifying descendants of John Carter Stark, a grandson of James Stark. The mutation at Marker 439 could have occurred with the birth of John Carter or with the birth of any direct males from John Carter to G. More descendants of John Carter Stark will need to be tested to confirm in which generation this mutation occurred.

The above analysis has shown all members of Group 2 share a common ancestor who lived within the last 20 generations (lived after the calendar year 1500). This suggests James, Richard and Archibald most likely have a common ancestor who lived previous to their generations (at least before 1668, the approximate birth year of Richard). Genealogical research will be required to identify this common ancestor of all members of Group 2.

Earlier genealogical research (undocumented) suggested Archibald and James were brothers and Dr. Richard Stark was their Uncle. In her publication entitled "The Family of General John Stark of New Hampshire, Jane Elizabeth Stark Maney¹ had this to say about "Other Stark Families in America."²

 "Some genealogists have suggested that James Stark may have been a brother of Archibald, our immigrant ancestor. In 1998, the Scottish genealogist, A. R. Bigwood, failed to find this relationship, in fact, indications are that they were not brothers.

Another early Stark family settled in York County, Virginia in the early 1700's, Richard Stark, who had been a physician in Glasgow, Scotland. He was a surgeon for a Scottish regiment of soldiers. He was born in Scotland before 1668, and is related to the James Stark and Archibald Stark families in some way. Some have suggested that he may have been an uncle of Archibald Stark." 

Using the genetic results of specific descendants in Group 2 having the H37 haplotype, a "Genetic Genealogy Report" has been prepared to address the issues presented in the Maney publication. 

^_

^________

Page 6

Part II: Genetic Genealogy Analysis of the Descendants of James Stark [1695-1754]

Introduction

The purpose of this dissertation is to examine the genetic and genealogical data of male descendants of James Stark [1695-1754] of Stafford County, Virginia. All the members of Subgroup 2C are genealogically known to be descendants of James Stark.

Subgroup 2C Genealogical Lineage

The Subgroup 2C Genealogical Lineage Table shows the direct male line from the Kit Number to James Stark. Reading from left to right, the earliest known Ancestor — James Stark — for each member of Subgroup 2C is shown in the Gen 9 column.

You will observe the participants in Subgroup 2C cover three generations (Gen 0, Gen 1, & Gen 2). The youngest are in Gen 0; the Parent Generation of Gen 0 would be Gen 1; and the grandparent Generation would be Gen 2. The second column from the left column shows the Panels tested for the member in that row and the number of mismatches in each Panel — relevant to the Ancestral Haplotype. Click on the Kit # for a more detailed report for each participant and his ancestors. Use your back button to return to this page.

The number of Unique Transmission Events (Births) in a generation can be found under the Gen number for that column. A UTE occurs when the Y-DNA of the Father is transmitted to his son. In each generation, the total births are the sum of all of the individuals listed in that Gen column. The total number of Births reported in Gen Columns 0 through 8 are presented in the Gen 9 column.

For this report, the Ancestral Haplotype derived from the Deductive Triangulation Method has be assigned to the genealogically determined common ancestor of all Members of Subgroup 2C, James Stark of Stafford County, Virginia. [Click HERE to review Appendix 1 entitled "Defining the Ancestral Haplotype of James Stark [1695-1754]."] Use Back Button to return to this text.

Subgroup 2C H12 Haplotype Life Expectancy Calculations  (Markers 1 thru 12) [See Appendix 2]

The genealogical data suggests there were 40 Births (UTE transmissions). H12 is a 12 marker haplotype over Panel 1. Member G has an allele value of 13 at DYS-439 while the most common allele value is 12, resulting in one mutation.

In the H12 Haplotype, there was one observed mutation. Solving equation (1) for K=1; B=40; and M=12:

R=K/B*M=1/40*12=1/480=(.00208) = the  mutation rate based on the total observed mutations in the H12 haplotypes

Solving for Appendix 2 equation (3):

(Average Life Expectancy of H12 Ancestral Haplotype) =1 / [1 - (1 - 0.00208)¹²]= approximately 40.52 generations.

If equation (3) is solved using the FTDNA average assumed mutation rate of (0.002) — the resulting number of generations will be 41.2. The member data will result in a H12 mutation rate that is approximately equal to the Standard mutation rate of 0.002.

Subgroup 2C H25 Haplotype Life Expectancy Calculations (Markers 1 thru 25) [See Appendix 2]

Member H was not tested beyond Marker 12. H has a total of 6 UTE transmissions that must be subtracted from the total UTE transmissions (40-6=34). Referring to the Genetic Results Table, the total mutation count will be 3.

Solving equation (1) for K=3; B=34;  and M=25:

R=3/34*25=3/850=(.00352)= the  mutation rate based on the total observed mismatches in the H25 haplotypes

Solving for Appendix 2 equation (3):

(Average Life Expectancy of H25 Ancestral Haplotype) =1 / [1 - (1 - 0.00352)²⁵]= approximately 11.85 generations

If the Standard mutation rate of (0.002) ---  assumed in the FTDNA literature --- is used in equation 3, the resulting number of generations will be 20.48. As more participants have been tested  the member mutation rate is greater than the Standard mutation rate of 0.002.

Subgroup 2C H37 Haplotype Life Expectancy Calculations (1 thru 37) [See Appendix 2]

Member H does not tested beyond Marker 12.  H has a total of 6 UTE transmissions that must be subtracted from the total UTE transmissions (40-6=34). Referring to the Genetic Results Table, the total mutation count will be 6.

Solving equation (1) for K=19; B=151; and M=37:

R=6/34*37=6/1258=(0.00476) = the  mutation rate based on the total observed mismatches in the H37 haplotypes

Solving for Appendix 2 equation (3):

(Average Life Expectancy of H37 Ancestral Haplotype) =1 / [1 - (1 - 0.00476)³⁷]= approximately 6.17 generations.

If the average assumed mutation rate of (0.002) --- assumed in the FTDNA literature --- is used in equation 3, the resulting number of generations will be 14. The member data will result in a H37 mutation rate that is considerably higher than the Standard mutation rate of 0.002.

Page 7

Appendix 1: Defining the Ancestral Haplotype of James Stark [1695-1754]

Introduction

Page 8

Appendix 2: Haplotype and DYS Marker Life Expectancy Calculations

Introduction

Mutation Rate: The rate at which a genetic marker or haplotype mutates or changes over time; expressed as a decimal value or a percentage. Commercial DNA testing laboratories use an average haplotype mutation rate most often given as 0.002, 0.003, and 0.004. Depending on the study, these average mutations for all of the markers in a haplotype applies to the general male population as a whole.

Surname projects having genealogical research of high quality with one common ancestor identified as the progenitor  —  can have average mutation rates that do not agree with the literature.  The number of participants and the genealogy of the descendants of their most recent common ancestor is of sufficient quality; it becomes possible to calculate the average mutation rate for various haplotypes and DYS Markers using Kerchner's equations. 

The mathematical model is based on an article by Charles F. Kerchner entitled, "An Overview and Discussion of Various DNA Mutation Rates and DNA Haplotype Mutation Rates." 

Haplotype Mutation Rates (Mathematical Model) 

Each Unique Transmission Event (birth) that occurred after the members common ancestor has been totaled in the Genealogical Lineage Table. For each birth, there can be two possible outcomes at each marker.  Either there was a mutation or there was not a mutation at the marker being examined. With this information, we can calculate an average mutation rate for each of the H12, H25, and H37 Ancestral Haplotypes using the equations of Charles F. Kerchner  as follows:

Mutation Rate = (Total haplotype mutations) divided by the (Total Births) times the (Total # markers in the haplotype)

Then for R= [Mutation Rate]; K= [Total haplotype mutations]; B= [Total Births]; and M=[Total # markers in the haplotype]

we have the equation: 

 

R=K/B*M    [Average Ancestral Haplotype Mutation Rate]    (1)

Given R, then:

Probability (new emerging haplotype) +Probability (old haplotype)=1

Probability (old haplotype) = (1-R)^M

Probability (new emerging haplotype) = 1 – (1-R)^M

Then if  P(N) = Probability (new emerging haplotype); we have the equation:

P(N)=1 - (1 - R)^M                (2)

A new haplotype can emerge at any time; but the above estimates the historically implied life expectancy of the Ancestral Haplotype, as defined by the genetic test results of Aaron's male descendants with the surname Stark --- and --- the genealogical data to define the number of generations. To convert this result to the number of generations, we simply divide  1 by P(N) obtaining the following equation:

(Average Life Expectancy of Ancestral Haplotype) = 1 / P(N)=1 / [1 - (1 -R)^M]  (Stated as the # of Generations)      (3)

This will be the estimated average life expectancy before a new haplotype will emerge within the perimeters of the genetic and genealogical data provided. Equation 3 has been used to calculate the average life expectancy of the H12, H25, and H37 Haplotypes.

Page 9

Appendix 3: Group 2 Selection Criteria

Page 10

Group 2 - Genetic Placement Criteria 12 Marker Comparisons

12 marker comparisons shall be made to the Group 2 12 Marker Modal Haplotype. Genetic comparisons to a present member having the 12 marker Modal Haplotype determine placement in Subgroup 2a, 2b, or 2c. If a new member passes the genetic comparison that follows but does not have a genealogical paper trail or an incomplete paper trail, the new members shall be placed in Subgroup 1d.  If a new member fails the genetic comparison, but has a genealogical paper trail suggesting Richard, Archibald or James was his ancestor, he shall be placed in Subgroup 1f.