Reconstructing Your Parentage and Ancestry

Every year in the United States about half a million paternity cases are performed proving or disproving whether an alleged father is the true parent of a child. Sometimes there is a court order to do this; at other times, it is sheerly for personal information. The determination of parentage is made based on a simple comparison of a small rock-hard number of genetic markers in the DNA fingerprint of the child and alleged father. Samples are extracted from a 30-second cheek swab and processed at any of an estimated 2,000 forensic labs across the country. The standard in place since about 1997 has been a set of 30-32 biallelic or double values each person carries on loci spread across their chromosomes, making for a virtually unique identification signature that reflects the equal DNA input of mother and father (and in fact all grandparents and all ancestors).

Often termed CODIS markers (standing for Combined DNA Identification System), these alleles or variations are the magic numbers underlying the popularity of paternity tests as well as the national passion for jailing or exonerating crime suspects. If a value is found in the DNA profile of the child and is not present in the two observed values of the alleged father on the same locus, this constitutes what is known in the paternity business as an exclusion:  the alleged father is almost certainly not the true father. Conversely, if all the alleged father’s values can be detected in the child’s on each location, one after another, that male is judged to be the child’s biological father to a 99.999% certainty. Paternity tests are simple math.

A famous paternity test involved proving who was the true father of the baby born to Anna Nicole Smith in 2006. After her death in early 2007, several men came forward claiming to be in father, including a European prince, Anna Nicole’s bodyguard and a convict who had been a former boyfriend. Larry Birkhead pressed his case. When the results came in, he was declared by Bahamian court to be the baby’s biological father. The child’s original birth certificate was amended to show this.

Can paternity testing be used in a reverse process to establish the identity of a father, given only the child’s DNA profile? No, but with enough DNA profiles available for comparison the missing member of a family group can be reconstructed by comparing alleles they must share, called obligate.  Doing so is a matter of logic and statistics, mostly just either-or, deductive logic.

I became interested in reconstructing a parent’s profile after many of DNA Consultants’ customers inquired if such a calculation or estimate was even possible. Some were adopted persons who had no recourse to testing their parents, some knew one parent but not the other, and some had no access to parents. They were either uninterested or unavailable. In a special category were females who were only-children with both parents deceased who wanted to know something about their father, but who could not take a Y-chromosome haplotyping test, as they did not carry a copy of their father’s male DNA. In this respect, autosomal DNA testing is the great equalizer.

My father, Lawden Henry Yates, died in 1978. My mother, Bessie Cooper Yates, lived to the advent of DNA tests, but I failed to obtain any sample from her before her death in 2006. I had siblings and half-siblings still living, however, so in 2010, I constructed a family group autosomal DNA study with the help of Crystal Wagner at Chromosomal Laboratories/Bode Technology. The results were very satisfying. This paper and blog post will serve as a report to those who are interested.

Step One
I was fortunate to have the participation of three half-sisters by my father, along with his second wife, their mother. Comparing mother and daughters I was able to verify the obligate alleles each daughter must have received from the mother.

Autosomal alleles are fixed in our genealogy, have little or no mutations (unlike YSTRs, which mutate from generation to generation, as do mitochondrial nucleotide positions, though more gradually over time) and derive from both parents equally by recombination at the moment of conception. They are copied and preserved without change in every cell of our bodies. The mother is responsible for half of the equation.  By a process of elimination the other number on each row of the lab report must represent the father’s contributions.  This method is completely logical and unequivocal. There can be no other answer to the problem. No studies suggest these pieces of our double helix DNA change significantly in transmission from one generation to the next or mutate over time in genealogies. Their values and patterns are strictly attributable to heredity.

Step Two
The father’s alleles are confirmed by a comparison with three children by his first wife, my mother.

Step Three
By the same watertight process we can now proceed to the mother’s reconstructed DNA profile. In it, we can expect to visualize the final piece of the puzzle, proceeding from the known to the unknown according to the immutable laws of autosomal DNA and genetic inheritance.

We have arrived at my mother Bessie Yates’ DNA profile by a multi-step process of extrapolating it using three of her children and three children by her husband’s second marriage, along with the test results of my half-sisters’ mother. Seven tested profiles yielded two reconstructed ones. In the process we have also recovered my deceased father’s DNA profile.

Separating Mother and Father’s Contributions to Ancestry
Having overcome these hurtles, I was most interested in the utility of the results. I felt confidant about the method. But what excited me most was to see how my own autosomal ancestry results might be respectively apportioned in my parents. For this, I ran a DNA Fingerprint Plus on them both. The findings were very satisfying to me personally, helping solve many questions I had always had about what ancestry I got from my father, what from my mother and what from both.

Let’s start with American Indian admixture. My own DNA Fingerprint Test, as well as percentage tests through another company, suggested a relatively large amount, perhaps one-quarter all told by various measures, but family tradition had placed Native American heritage solely on my mother’s side. To be sure, my mother gave me a Native American mitochondrial haplotype, indicating a female line going back to a Cherokee woman in Georgia, traced as far back in records as 1790. Extensive genealogy research showed, however, that my father’s great-grandmother was also a Cherokee with the surname Thomas from North Carolina. What did the new autosomal DNA profiles say?

On a rough measure, I have received a “double dose” of Native American II, a marker co-relating with 80% of 24 tested American Indian populations in the atDNA 4.0 database. (Two siblings and one half-sibling received only single doses.) This seemed to indicate that I had some degree Native American (not possible to say how much) from both parents. True enough apparently, judging from the top world matches for my mother and father. I give here the top ten for comparison.

My mother’s Native American population matches were slightly higher and more numerous than my father’s, including more peoples like the Chukchi and Mongols, but my father’s were not inconsiderable in their own right. Here’s how their two megapopulation rankings look:

These results confirmed that my father did have some Native American, although evidently not as much. They also suggested that although both bore about the same mixture of European and Native American ancestry (including high matches to Melungeon), my mother had a more pronounced Native American cast, her highest match being to North Asian, one of the supposed Asiatic feeder populations of Native Americans, whereas my father’s top match was European American. Based on profile frequencies, my father was five times more likely to be European American than American Indian if subjected to forensic profiling, whereas my mother was 18 times more likely to come out as a Siberian Native than Northern European. Sometimes, it seems, exotic ancestry rises to the top. My overall conclusion was that my mother probably had 3/8 and my father 1/8 Native American heritage, which corresponds to their proved genealogies.In my own profile, combining those of my parents, here are my megapopulation results:
Self (Donald N. Yates)

According to these frequencies, my mother and father’s Native American ancestry reinforced each other in me to make my top four matches Native American (or Siberian-Mongol-Turkic), so that I am about twice as likely to be graded into the Native American category by population statistics than the European. Similar conclusions emerged from my siblings’ tests, and a diminished presence of Native American indicators was confirmed in my half-siblings, although their mother seemed to evince some Native American as well as my father, the shared parent. All participants in this study had grandparents born in North Alabama.Further observations are possible. For instance, I was surprised to see a large indication of Jewish ancestry in my father’s profile. Genealogy confirms as much, as the family surname is Hebrew (an anagram of Ger Tzedek similar to Katz, Kohen Tzedek). The emigrant Yates figure was reportedly an English Jew in seventeenth-century Virginia. My mother also showed Jewish ancestry. Both parents matched Melungeons, an Appalachian ethnic type suspected to have Sephardic Jewish forebears. My father’s family included uncles named Josephus, Manaen, Irbin, Azariah, Lazarus and Sherith—apparently his Middle Eastern matches were truthful to a partial Muslim background. My mother’s mother was named Palestine, and the names Isaac and Jacob were ubiquitous in her family tree. But neither side of the family claimed any Jewish heritage. It was left to autosomal DNA to reveal that hidden inheritance.Although never performed before to my knowledge, this method of reconstructing autosomal profiles can be useful to others seeking to recover unavailable relatives’ genetic fingerprints and to separate parents’ contributions to their children’s ethnic and ancestral stories. Since it is based on immutable markers in DNA it rests on more solid ground than Y chromosome alleles or mitochondrial mutations. The challenge in exploiting the method is to have enough subjects in your family group study. In my case, I was fortunate to have a prolific father with six living children. I would like to conclude by thanking all my siblings, half-sisters and my father’s widow. Their participation made it possible to present a true first in DNA genealogy.