The Rattlesnake Ridge Genealogy Club
April 30, 2016
Presentation on DNA Genealogy
John W. Grace
Topics Covered
- FamilyTreeDNA resources
- Developing a Research Plan
- Your “Biological Surname(s)”
- Contacting your “biological cousins”
- A suggested DNA testing protocol
- My unexpected DNA test results
- What are STR’s?
- What are Autosomal tests?
- A Hypothetical STR Inheritance Family Tree
Family Tree DNA Resources
Most of the information we gather using traditional research methods is free (Ancestry.com subscriptions not withstanding). The cost of DNA testing is not trivial: Individual tests can cost from $99 to almost $600. So, it's best to have a plan and to set clear DNA test investment priorities. Impulsive DNA test purchases can be costly and serve little purpose.
And, be prepared to subsidize the testing of prospective cousins; and don’t expect to find a lot of
DNA matches immediately. Although the number of individuals who have been testing now numbers in the hundreds of thousands, you will find relatively few of them enrolled in surname projects, and fewer still will be a close match to you. Realistically, you’ll be lucky to find more than one or two. As time goes on, your odds will no doubt improve.
About The Family Tree DNA Database
Our databases are the most comprehensive in the field of Genetic Genealogy. As of April 9, 2016, the Family Tree DNA database has 783,158 records. Total numbers include transfers from the Genographic Project and resellers in Europe and Middle East. We also have:
8,810 surname projects
413,918 unique surnames
564,131 Y-DNA records in the database
251,200 25-marker records in the database
230,375 37-marker records in the database
116,532 67-marker records in the database
219,027 mtDNA records in the database
72,829 FGS records in the database
Family
Tree DNA Tests
mtDNA test: $199
Y-DNA 37 marker test: $169
Y-DNA 67 marker test: $268
Y-DNA 111 marker test: $359
Family Finder autosonmal test: $99
Comprehensive Genome (Y-DNA 67 + Family Finder + mtDNA test): $566
Other Resources
National Geographic's Genographic Project test kit: $148.95
AncestryDNA test kit: $99
23andme test kit $199
A Plan
This is one approach, you may well find another that works better for you.
- Do a 67 marker test ($268) on yourself . FamilyTreeDNA will notify you of your matches. If you have any matches at the 67 marker level, contact the ones that are within a genetic distance of 2 or less. If your surname matches theirs, join the project for your surname. If your DNA matches multiple testees with a different surname from yours, join their surname project.
Confer with those with other researchers to see if you can determine your correct biological line of descent. If you’re lucky, your biological descent will match your surname descent and you can use your test results to overcome brick walls and find new cousins. Simple. If genetic profile matches those of a different surname (as mine did), then things get more complicated.
Be sure to check for matches at the 37 marker level. Many researchers do not order a 67 marker test, and if you don’t check for matches at the 37 marker level you may well miss several matches that will turn out to be just as important to you as the ones you find with the higher resolution tests. Matches at less than the 37 marker level seldom prove valuable.
This would be a good point to create a GedMatch.com account. Once you have established your account, login into your FamilyTreeDNA account (or other, as appropriate), then download your 37 marker concatenated DNA results database to your PC. Log into your GedMatch.com account and upload your concatenated DNA results file. It will take 24 hours for the batch update process to complete.
When things get complicated.
If your biological line is one that you were unaware of, then you can look forward to starting a whole new family tree from scratch, researching what I will call your “biological surname”. Before you begin, see if you can determine when the surname anomaly occurred. There are multiple approaches to this problem.
- Look at your family’s census entries, and look for individuals/families sharing your biological surname residing in the same area.
- Look for your biological surname among the marriages of your cousins, 2nd counsins,3rd cousins, etc., especially those who were living in the geographic area of your inherited surname kinfolk. Don’t forget cemeteries, church membership rosters, club membership rosters, etc. Think like a detective.
Be prepared to “think different”
Traditional genealogical research is “surname centric”, and it is very difficult for any of us to escape that mindset. To use genetic results effectively, we must look upon ourselves as having a “biological identity”, distinct from (and complementary to) our family identity as defined by our surname.
Never forget the limitations of research based upon our surname:
The surname we carry may be the result of a formal or informal adoption. It may be the result of misattribution of parentage. It may be the result of one person’s willful desire to break themselves free of their old identity and to “start a new life” under an assumed or “Americanized” name. Lastly, as we all know very well, people of good intention can come to have their surname expressed in a variety of ways. The diverse local versions of Maddix, Vanlandingham, Fultz, Rayburn, Gilliam and P‘Simer are all testaments to that fact. Farrier, Ferrier, Ferrer, Ferrar, Farrer, Farrah, Farrey, Farrow, Faro, Pharrow, Pharoah, Varah, Varey, Varrow and Vairow all share the same surname root!
We must also keep in mind that inherited surnames are a relatively new phenomena, particularly among those of us not closely linked to medieval royalty. The Normans introduced surnames to England after 1066. The Irish and the Scotts only began adopting inherited surnames after their union with England in 1536.
All said, it is little wonder that many of us get almost weekly notifications from FamilyTreeDNA of new “genetic cousins” who never seem to share our last name ;)
If you can use traditional genealogical methods to trace your ancestry back to 1800, you are in all probability, nearly half way back to the beginning of the existence of your surname!
Genetic test results are the only means that we have available to us to establish our true biological relationships with our fellow human beings.
Make Contact
If you find some candidate lines that share your biological surname and get enough detail so that you can locate the living descendants who have the highest probability of being related to you. You may find only one, more likely you will find a dozen or more. You’ll probably be able to isolate your candidate lines to one to four or five. Plan on testing at least one living individual in each of the candidate lines.
Use the internet and any helpful contacts you may come across to locate your candidate cousins. The internet is great for finding telephone numbers, especially if you can narrow the residence area of the family your are researching. Many families stay in the same state or county for long periods of time. Hopefully you have middle names and the names of relatives to narrow your field of candidates.
Now call or write to you candidate cousins. Try to resist pressing them for person information until you have established a rapport. Give them as much detail about yourself and your reason for contacting them as you feel comfortable with. The more forthcoming you are, the more the other person will feel comfortable discussing their family with you.
Many people know nothing about DNA testing, so you may have to do some explaining before you ask them to do a DNA test for you. If they are already actively interested in genealogical research, you might suggest splitting the cost of the test, more probably, you will have to offer to cover the cost. Your offer will be seen as an indication of your seriousness and, from my limited experience, re-assure them that you’re not trying to take advantage of them for some ulterior motive.
Your first contact may well not agree to the test. If they don’t, thank them for their time and ask them if any male members of the family have an interest in genealogy, and get their telephone number. Again, be a friendly detective.
Additional Testing
Order a 37 marker test ($169) for your candidate cousin. Be prepared to wait. Test turn around can take a 2-3 months, and 6 weeks is typical. This level of testing will tell you if you are a close genetic match. If the first one that you test is a non-match, find another candidate in the same family line a repeat the “Make Contact” process.
If you are a close match (genetic distance of 0 or 1, upgrade their test to the 67 marker test ($99 upgrade - $79 on sale) level. FamilyTreeDNA will use the original test kit sample for the upgrade.
If you suspect that a different line of descent might produce an even closer match, repeat the “Make Contact” and “Get the test Done” for the second (or more) candidate line. If you have multiple matches that are very close, you will want to upgrade the relevant kits (including your own of course) to the 111 marker level ($129 upgrade - $109 on sale).
An autosomal test is said to be the best way to zero in on your biological family group.
A DNA Test Protocol
I suggest your test protocol be some variation of:
- A 67 or 111 marker DNA test for yourself
- Upload of your DNA results and ancestral GEDCOM file to GedMatch.com
- 37 marker tests for identifiable living likely candidate cousins
- Full 67 or 111 marker tests for the “finalist” candidate cousins (hopefully less than 4)
- Autosomal tests for yourself and your closest cousin matches (to isolate your MRCA)
My Search
Screenshots from the Grace Surname Project page at FamilyTreeDNA
Below is a screenshot from the Wingarner/Winegartner Project page
Although my surname is Grace, my DNA is evidence that I am a descendant of a Winegardener 3rd great grandfather. I dedicate a portion of each week attempting to track down exactly who he was.
I frequently correspond with the Winegardener researcher who is a genetic distance of 1 from me at the 67 marker level. Gary has generously provided me with leads to over a dozen living individuals who may be in my line of descent. I am in the process of contacting them, with hopes that they will allow me to obtain a DNA test sample from them.
My Family Tree DNA Y-DNA TiP Report
In comparing Y-DNA 37 marker results, the probability that Carl Clemens and John Wesley Grace shared a common ancestor within the last...
Genetic Distance Comparison Chart
Carl Clemens is a genetic distance of 4 from me at the 37 marker level. I used 30 year generations to estimate the time frame involved.
Generations Percentage
4 4.03% - 1 to 120 years
8 27.62% - 240 years
12 58.67% - 360 years
16 80.81% - 480 years
20 92.30% - 600 years
24 97.22% - 720 years (beyond 600-720 years pre-dates surname usage)
Once one goes beyond a genetic distance of 4, the number of biological cousins we have that do not share our surname increases exponentially, and any linkage to modern surnames becomes largely coincidental.
Always keep in mind that
percentages only express probabilities. When a relationship has a 9 chances out
of 10 of occuring, it still has a 1 chance in 10 of NOT occuring.
What are STR’s?
Highlights from the link above:
Y-STR stands for the Y chromosome and STR is short tandem repeat. The tandem repeat refers to a series of DNA that is repeated a certain number of times. The number of times this is repeated varies among men and gives us a value for each Y-STR marker we test. A marker is simply the generic term we use when referring to a STR. The results looks very simple:
DYS 393 = 13
Translation: the marker called DYS 393 has a value of 13, which means the sequence (bases in a particular order such as: TATT) is repeated 13 times:
TATTTATTTATTTATTTATTTATTTATTTATTTATTTATTTATTTATTTATT
A set of these marker values gives us a unique signature, also known as a haplotype. A Y-DNA 37-marker test will give a list of 37 markers and their values. When looking at a list of Y-DNA results at first they appear to be all fairly similar. But a closer look will yield patterns.
Here is a screenshot of part of the Wheaton surname project results page at FTDNA
The first four columns are self explanatory (Kit Number, Name etc---we will get to Haplogroups in the next lesson). The fifth and subsequent columns labeled DYS363, DYS390 etc. refer to the STR marker's name. If you look closely you will see that the results are clustered into Groups denoted by a horizontal colored row and a group description. If we look at the two kits in Group "C" we can see that the values are identical except for the first value at the CDY marker (slide your horizontal scroll bar all the way to the right if you cannot see). So for the CDY value the Kit #232717 has a different value (mutation) than the one below it.
A mutation is simply a change in value due to a replication error. So in our example instead of 39 repeats Kit# 232717 dropped a repeat so all of his direct male descendants will carry a value of 38 at CDYa. Lest you think replication errors are bad they are the way in which we can trace how closely two men are related. Some markers are more prone to mutations than others. You will note that some of the columns are blue and some are burgundy. The faster mutating markers are shown in burgundy. So differences in values at these markers are not as significant as those at the slowest mutating markers (blue). Mutations do not happen very frequently but they do happen randomly from time to time. They also happen with slightly higher frequency the older the father is at the birth of his son. You will note that values are shaded light pink, dark pink, light blue and dark blue. If the value is one below the modal (average) it will be light blue, more than one below is darker blue. If the value is one above the modal it will be light pink, more than one above dark pink.
Let's take a closer look at Group "B" (above). All of the men in group B are related in a "genealogical time frame." This genealogical time frame refers to the period since surname adoption, roughly since about the 1400-1500's. You will note that several of the men do not bear the surname Wheaton. Remember when I said DNA doesn't care what we want it just reports what is there. We do not know which surname came first but what we do know is that all of these men are descended from a common ancestor on the patrilineal (father's father's father's line). More on how this can happen on the page Dealing With The Unexpected Result.
This screenshot shows you what a group of men who are not related on the "Y" looks like. Note that the haplogroups (I1, R1b1a2, G) are different and the values differ greatly even though some values at some markers are the same. These are men who participate in the Wheaton project via their atDNA but they do not bear the surname Wheaton and therefore do not match any of our Wheaton profiles.
Note that there are literally millions of STR’s in human DNA. Not all of them are useful for genealogical purposes. The ones that are “Goldie Locks” STR’s mutate over time, but not too frequently and not too rarely. They must also be the ones which can be analyzed on an affordable DNA test.
Autosomal tests
Contained in the nucleus of each cell are twenty-three pairs of chromosomes.
Twenty-two of these matched pairs of chromosomes are called
"autosomes," while the 23rd pair determines your sex (male or
female). Autosomal DNA is inherited from both parents, and includes random
contributions from their parents, grandparents, and so on. Therefore, your autosomes essentially contain a complete genetic record, with all branches of your ancestry at some point contributing a piece of your autosomal DNA.
Autosomal DNA tests can be used to search for relative connections along any branch of your family tree. Unless the connection is so far back that the shared DNA has essentially been eliminated through too many generations of recombination, any autosomal match between two individuals indicates a possible genetic connection. There is nothing in this test that will tell you which branch of your family the match is on, however.
Therefore, having your parents, grandparents, cousins, and other family members tested will help you to narrow down potential matches.
Available To:
Everyone (males and females)
How Autosomal DNA Testing Works:
For each of your twenty-two pairs of autosomal chromosomes, you received one from your mother and one from your father. Before they passed these chromosomes down to you, the contents were randomly jumbled in a process called "recombination" (this is why you and your siblings are all a little different from each other). Your parents, in turn, received their chromosomes from their parents (your grandparents). Your autosomal DNA, therefore, contains random bits of DNA from your great-grandparents, great-great grandparents, and so on.
Close relatives will share large fragments of DNA from a common ancestor. Connections arising from more distant relatives will result in smaller fragments of shared DNA. The smaller the fragment of shared autosomal DNA, generally the further back the connection in your family tree. Even these tiny segments of shared DNA can potentially hold a clue, however! The way in which your individual DNA has recombined through the generations also means that you may no longer carry DNA from a particular ancestor. Distant relatives often share no genetic material at all, although it is also possible to match an individual through a very distant ancestor.
How Accurate is Autosomal DNA Testing?:
The average amount of autosomal DNA shared with a relative decreases with each successive generation. Percentages are also approximate - for example a sibling may share anywhere from 47–52% of their DNA in common.
•50% (parents and siblings)
•25% (grandparents, aunts/uncles, half-siblings, double first-cousins)
•12.5% (first cousins)
•6.25% (first cousins, once removed)
•3.125 (second cousins, first cousins twice removed)
•0.781% (third cousins)
•0.195% (fourth cousins)
The chance that an autosomal DNA test will accurately detect a relative decreases with the distance of the relationship. For example, most autosomal DNA ancestry tests predict an accuracy rate of 90–98% when detecting a match with a 3rd cousin, but around a 45–50% chance of detecting a match with a fourth cousin. Depending on the DNA recombination, however, an autosomal test may sometimes accurately detect more distant cousins (fifth cousins and beyond). Double descent from a common distant ancestor (e.g. marriage of second cousins) may potentially increase the chance of a match.
What Autosomal DNA Test Should I Have Done?:
A number of different companies are offering autosomal DNA tests, with a few offering databases to help you use your results to connect with other potential relatives. Three of the largest include (alphabetical order):
• AncestryDNA from Ancestry.com
• Family Finder from FamilyTreeDNA.com
• Relative Finder from 23andme.com
As stated above, when your autosomal DNA is analyzed and the results provided to you, half of the DNA sequences you will be given came to you from your mother, and half from your father. What the test cannot tell you is, which half is which. That part is left for you to determine. The good news is, there is a well established procedure for doing this.
From the Family Tree DNA Message Board:
“With phasing you always start with the easiest portions of the DNA that can be phased and then you branch out from there. In general with a two parent/one child trio you can phase about 94% of the autosomal data. Since siblings share roughly 50% of their DNA, the data for about 47% (.5 x
.94) of the other two siblings' data should be able to be accurately phased. Since a paternal/maternal aunt shares about 25% of her DNA with her nieces and nephews this would mean that you should be able to readily phase about 23.5% (.25 x .94) of the aunt's DNA. A higher percentage of the aunt's data could be phased in situations where the aunt and the other two siblings are homozygous for specific SNPs in large HIRs that the other two siblings share in common with the aunt. In theory, 3 siblings should collectively have 87.5% of the autosomal DNA of their parents. The aunt will share about 50% of her DNA
with her deceased sibling. Thus
the aunt should share about 43.75% (.875 x .5) of her DNA with the three
siblings. This would be the maximum amount that could be phased. Thus, you
should be able to infer a minimum of 23.5% the genetics of the deceased parents
and a maximum of 43.75%. I think that the actual percentage of the deceased
parents DNA that could be accurately inferred is about 30%. the three
siblings.” Posted by Tim Janzen
A Hypothetical STR Inheritance Family Tree
Really interesting DNA link:
http://www.marres.education/far-forbears.htm
Autosomal DNA inheritance odds:
Chances of no DNA from one grandparent: 1 in 8.4 million
Chances of no DNA from one great grandparent: 1 in 4096 (assumes 12 passed down in last generation)
Chances of no DNA from one great, great grandparent: 1 in 64 (assumes 6 passed down in last generation)
Chances of no DNA from one great, great, great grandparent: 1 in 8 (assumes 3 passed down in the last generation)
GEDMATCH
Uploading your concatenated FTDNA test results to Gedmatch
Note: before uploading your DNA results to GEDmatch, you must have purchased the FTDNA Family Finder test.
- login to your FTDNA account
- Write down your Y haplogroup ID (it will be something like "RM296")
- On your FTDNA account screen, you will see a small link that says "Download Raw Data".
Click on that link.
- On the next page, click on the link that says "Build 37 X Raw Data Concatenated"
- Save the file to your hard drive
- After the file is on your hard drive, login to GEDmatch (create a new account if you don't already have one). https://www.gedmatch.com/login1.php
- On the right side of your GEDMatch account screen, you'll see links under "Raw DNA File Uploads". Click on the link that says "FTDNA Concatenated DNA file"
- On the next screen ("GEDmatch Data upload") you'll see a box that says "Name of Donor". Enter your name there, and an alias name in the next box.
- Fill in the other boxes, including the one for your Y haplogroup
- Click on the link that says "Upload".
Read the information on each of the screens as you move along. It will take 24 hours for your GEDmatch results to become available.
- After 24 hours, login to GEDMatch
- On your account screen, under "Analyze your data", click on the "One to many" matches.
- Explore the other links too
GEDMATCH
These matches are calculated from DNA raw data files submitted by users. Results are based on the default parameters for the GEDmatch Autosomal comparison software. These results are for comparison only and will likely be somewhat different than those shown by your testing company.
Default parameters (and results) may change periodically based on feedback from users, and from new research results.
Click on 'L' for list of matches for that person (May take up to 5 minutes to process)
Click on 'A' for details on Autosomal comparison.
Click on 'X' for details on X-DNA comparison.
Select 2 or more from 'Select' column, and click this button for additional display and processing options.
Intensity of GREEN background indicates how recent (within last 30 days) a match is.
To qualify as a 'match' in the genealogical time frame, results must have a largest Autosomal segment that has at least 700 SNPs and be at least 7 cM. It must have BOTH. Results with the largest segment less than 7 cM are highlighted in pink.
In general, the results shown below use thresholds LESS than 7cM / 700 SNPs.
PLEASE verify any result shown on this list with the one-to-one comparison tool before assuming any match is real.
To check the number of SNPs, click on the 'A' on the same line to view the one-to-one comparison detail.
Please DO NOT send emails to anyone on this list without first using the one-to-one utility to verify that it is a legitimate match. DO NOT create mass mailing lists from these results.
One to one match report
ChrStart LocationEnd LocationCentimorgans (cM)SNPs32,723,0764,793,1277.78885101,162,958154,490,82750.611,806Largest segment = 50.6 cM
Total of segments > 7 cM = 58.2 cM
Estimated number of generations to MRCA = 4.0
657164 SNPs used for this comparison.