Abstract from American Society of Human Genetics meeting session

The GenoChip: a new tool for genetic anthropology.


S. Wells1, E. Greenspan2, S. Staats2, T. Krahn2, C. Tyler-Smith3, Y. Xue3, S. Tofanelli4, P. Francalacci5, F. Cucca6, L. Pagani7, L. Jin8, H. Li8, T. G. Schurr9, J. B. Gaieski9, C. Melendez9, M. G. Vilar9, A. C. Owings9, R. Gomez10, R. Fujita11, F. Santos12, D. Comas13, O. Balanovsky14, E. Balanovska14, P. Zalloua15, H. Soodyall16, R. Pitchappan17, G. Arun Kumar17, M. F. Hammer18, B. Greenspan2, E. Elhaik19

1) Mission Programs, National Geographic Society, Washington, DC; 2) Family Tree DNA, Houston, TX; 3) Wellcome Trust Sanger Institute, Hinxton, UK; 4) University of Pisa, Italy; 5) University of Sassari, Italy; 6) National Research Council, Monserrato, Italy; 7) University of Cambridge, UK; 8) Fudan University, Shanghai, China; 9) University of Pennsylvania, Philadelphia, PA; 10) CINVESTAV, Mexico City, Mexico; 11) University of San Martin de Porres, Lima, Peru; 12) Federal University of Minas Gerais (UFMG), Belo Horizonte, Brazil; 13) Pompeu Fabra University, Barcelona, Spain; 14) Russian Academy of Medical Sciences, Moscow, Russia; 15) Lebanese American University, Byblos, Lebanon; 16) University of the Witwatersrand, Johannesburg, South Africa; 17) Chettinad Academy of Research and Education, Chennai, India; 18) University of Arizona, Tucson, AZ; 19) Johns Hopkins University, Baltimore, MD.

Background: The Genographic Project is an international effort aimed at charting human history using genetic data. The project is non-profit and non-medical, and through the sale of its public participation kits it supports cultural preservation efforts in indigenous and traditional communities. To extend our knowledge of the human journey, interbreeding with ancient hominins, and modern human demographic history, we designed a genotyping chip optimized for genetic anthropology research. Methods: Our goal was to design, produce, and validate a SNP array dedicated to genetic anthropology. The GenoChip is an Illumina HD iSelect genotyping bead array with over 130,000 highly informative autosomal and X-chromosomal SNPs ascertained from over 450 worldwide populations, ~13,000 Y-chromosomal SNPs, and ~3,000 mtDNA SNPs. To determine the extent of gene flow from archaic hominins to modern humans, we included over 25,000 SNPs from candidate regions of interbreeding between extinct hominins (Neanderthal and Denisovan) and modern humans. To avoid any inadvertent medical testing we filtered out all SNPs that have known or suspected health or functional associations. We validated the chip by genotyping over 1,000 samples from 1000 Genomes, Family Tree DNA, and Genographic Project populations. Results: The concordance between the GenoChip and the 1000 Genomes data was over 99.5%. The GenoChip has a SNP density of approximately (1/100,000) bases over 92% of the human genome and is highly compatible with Illumina and Affymetrix commercial platforms. The ~10,000 novel Y SNPs included on the chip have greatly refined our understanding of the Y-chromosome phylogenetic tree. By including Y and mtDNA SNPs on an unprecedented scale, the GenoChip is able to delineate extremely detailed human migratory paths. The autosomal and X-chromosomal markers included on the GenoChip have revealed novel patterns of ancestry that shed a detailed new light on human history. Interbreeding analysis with extinct hominids confirmed some previous reports and allowed us to describe the modern geographical distribution of these markers in detail. Conclusions: The GenoChip is the first genotyping chip completely dedicated to genetic anthropology with no known medically relevant markers. We anticipate that the large-scale application of the GenoChip using the Genographic Project’s diverse sample collection will provide new insights into genetic anthropology and human history.

You may contact the first author (during and after the meeting) at spwells@ngs.org

X Chromosome matches from 23andme

This chart is of my X Chromosome matches from Relative Finder and Ancestry Finder at 23andMe. The length of the line corresponds to the length of the match on the X chromosome. Because I am a male, I got one X chromosome from my mother.  She got an amalgam of the X chromosomes from her father and mother. I can only verify my relationship to three of the people on this chart.

SKD is my multiple 1st cousin and the relationship is through our mothers, who were sisters.  My 1st cousin and I share 6 of 8 great grandparents. We match at two locations on the X: 1-12, and 91-140.

LB and CF are 3rd cousins 1 remove and 2 removes, a mother and daughter. My 1st cousin and I share 4 of 16 GGreat grandparents with them, Swain, Ball, Kidd, and Stephens.  However, my 1st cousin does not match them or me at these locations on the X chromosome, but does match them on other chromosomes as I also do.

With HP, the link may be through the Hays/Hayes family of Overton and Fentress Co., Tennessee. HPs X line is from the same area. The mother of Maliza Stephens was Susan Hayes, and she and her husband, Zorababel Stephens, lived in Overton/Fentress Co., TN. This would subsume the links to LB and CF.

With three others, GA, JO, and CO, a mother and two of her sons, the relationship is unknown. My 1st cousin does share this match with them. We thought the relationship was on the Strunk/Strunck line, but that is not a direct line for them.

It is important to remember that a different cutoff is used for calling a match on the X chromosome between males, a male and a female, and between females:

Males: 200 SNPs, 1cM;
Male to Female: 600 SNPs, 6cM;
Female to Female: 1200 SNPs, 6cM.

This probably explains some match I have with females which do not show as a match for my female 1st cousin.

I am hoping to get some more 1st cousin matches to help to isolate matches to particular lines.

Here is a link to my pedigree charts:
Steven C Perkins 13 Generation Pedigree Charts 

X chromosome inheritance at 23andMe through Relative Finder and Ancestry Finder

The link above goes to a page that has charts for determining which ancestors contributed to your X chromosome. For a male, only his mother contributes an X chromosome. For a female both her mother and her father contribute X chromosomes. Parts of those chromosomes will be inactivated but the X can still be used for genealogical purposes.

I have 24 people who match me on the X chromosome at 23andMe through Relative Finder and Ancestry Finder. Neither FTDNA nor SMGF have published their X chromosome databases. GEDMATCH allows FTDNA customers to match their X chromosomes to other GEDMATCH participants. I haven't heard of anyone else having as many matches, but I am sure there must be someone who does. At this time I can only confirm three of the matches with one 1st cousin and with a 3rd cousin once removed and her daughter. I share a surname in the 1600s in Germany with a mother and her two sons. But since that surname is not contributing to the X chromosome before 1830, the assumption is that there is a closer link than that surname.

There are 5 and 6 generation charts for both males and females on the page. Below are the people, with their ancestor chart number, who would be on my 6 Generation Chart on the linked page with the percentage of X chromosome contribution expected from each.

1, myself
-------------
2 Denval Perkins 0%
3 Mary Ruth Ball 100%
------------
6 George Matt Ball 50%
7 Rosa Genetta Swain 50%
------------
12 Peter Coleman Ball 0%
13 Frances Strunk 50%
14 Thomas J Swain 25%
15 Elizabeth Kidd 25%
------------
26 Manoah Strunk 25%
27 Susana Cortina Davis 25%
28 Jesse D. Swain 0%
29 Elizabeth Ball 25%
30 John Kidd 12.5%
31 Maliza Stephens 12.5%
-------------
52 Abraham Strunk 0%
53 Abigail Pennington 25%
54 John Davis 12.5%
55 Frances Creekmore 12.5%
58 William Ball 12.5%
59 Charlotte May 12.5%
60 Elias Kidd 0%
61 Margaret Bagley 12.5%
62 Zorababel Stephens 6.25%
63 Susan Hayes 6.25%

Percentages are averages and usually can range from 0-100.

Having this chart filled out based on your sex can assist in identifying your X chromosome matches.

Comparison of 23andme Relative Finder, Ancestry Finder and FTDNA Family Finder results

I'll repeat a post I made at 23andme on comparing success rates at both:

======================================

Colonial American ancestry: Jamestown, Boston, Cambridge, Hartford, New Haven Colony, New Amsterdam, New Sweden, New Jersey, Pennsylvania. Then spread out to North Carolina, South Carolina, both pre-Rev War; then Tennessee and Kentucky, post-Rev War.

• at 23andme I have 1050 RF matches and have confirmed 7 of those.
• at FTDNA I have 175 FF matches and have confirmed 10 of those.
• Two of my confirmed matches have tested both 23andme RF and FTDNA FF.
• I have 20 separate X Chromosome matches in RF and AF. That counts repeated people only once.
• I have no immediate matches at FTDNA. I have 54 close matches and 121 speculative matches. My closest confirmed matches are two 4th cousins.
• I have two known 1st cousins at 23andme and two unresponsive 2nd cousins. My closest confirmed match at 23andme is with a 3rd cousin once removed and her daughter.

=======================

If you haven't done any tests I recommend starting at 23andme and then, when FTDNA opens the program, transferring your results to FTDNA. That way you get genealogical matches from both companies, you get mtDNA haplogroup and if a male, Y DNA haplogroup, and X chromosome matches at 23andme plus health info.

At FTDNA you get genealogy matches and when they get it ready, X chromosome matches. You have to pay separately for the Y DNA haplotype and haplogroup and the same for mtDNA. So from a cost perspective it is lower cost to start at 23andme and transfer to FTDNA UNLESS you also want detailed Y DNA /mtDNA readings.

Genetic Genealogy testing: Sorenson Molecular Genealogy Foundation

We often focus on the commercial DNA testing companies and forget about the non-profit Sorenson Molecular Genealogy Foundation, http://smgf.org/ They were the first organization to use DNA testing for genealogy. They began with testing of blood and then moved on to testing with a mouth-wash method. One of their goals was to build a collection 100,000 samples. This was reached a few years ago. These samples have been tested for Y chromosome DNA, mitochondrial mtDNA, and autosomal DNA from Chromosomes 1 through 22 and the X chromosome. At this time the Y DNA and mtDNA results are available for searching by entering your results or by using a surname to search the genealogical charts the participants sent with their samples. The autosomal database has not yet been released for searching.

Did I mention the test is free of charge?

One drawback to the testing is that there is no guarantee that your results will make it into the public databases where you can retrieve them at no charge. However, you can get your results from GeneTree, http://www.genetree.com for a nominal charge.

At this time SMGF is accepting samples from the following populations:

Other Areas of Interest
Reaching the 100,000 participants milestone was just one of the objectives of our project. We are still working to enhance our dataset by actively seeking participants for our database with ancestry from a number of countries and lineages (listed below). If you have genealogical information linking you to one of these areas or have connections that could facilitate collections in these areas please contact SMGF at info@smgf.org.

• Ireland
• Japan
• Sweden
• France
• Germany
• Belgium
• All Eastern European countries
• Lebanon
• Syria
• Jordan
• Egypt
• North Africa
• Switzerland
• Denmark
• Spain
• Portugal

Family Finder, Relative Finder, deCODEme, SMGF

In my capacity of running 9 DNA studies and advising people on testing, I have been looking at the information on Family Finder from FTDNA in comparison to 23andMe's Relative Finder Ancestry Edition.

Family Finder is only looking at the non-sex chromosomes and it does not include mtDNA testing. If we include the minimal testing of Y dna and/or mtDNA at the Genographic Project (to get the lowest price) a female would undercut the price of the 23andMe Relative Finder Ancestry Edition test, but would still not have the X chromosome tested, and a male would pay more than Relative Finder Ancestry Edition and still not have the X chromsome tested.

My ancestry is mainly Colonial American from the British Isles with some French, German, and Norwegian. My experience with Relative Finder is that of the 30,000 plus people in the project, I have matches with 307. I have sent out 106 sharing requests with 5 declines, 69 contacts accepted, and 32 contacts with sharing. Of the 32 sharing I have found the common ancestor for 6 of them. 3 of them share on the X chromosome. Hopefully as the database grows I'll have more matches and shares.

I think the companies should require a 5 or 6 generation pedigree chart from the people testing which would be shared with their matches. I think that would really help with the process of finding common ancestors.

This also brings up the question of what happened to the Sorenson Molecular Genealogy Foundation's plans to release an autosomal database which would be linked to the pedigrees they collected. Does anyone have any information from SMGF?

While I am getting the Family Finder testing, and have done the 23andme complete edition, exchanged with deCODEme, and the SMGF testing, I am somewhat disappointed in the results so far, and will have to really see some improvements before I can recommend these tests to any but the most dedicated genetic genealogists.

Autosomal connections

Sorenson Molecular Genealogy Foundation began a project to test the DNA of 100,000 people. They have recently reached that goal and are now testing only selected individuals. SMGF has made available the Y DNA and mtDNA results from those tests. You can search those results using your own DNA results from another company, or by searching for the name of a person on the pedigree charts submitted by the people tested. In addition to the Y DNA and mtDNA, SMGF is supposed to release a database of autosomal marker results to the public. In this regard it seems they are about to be beaten to press by 23andMe.

A video on autosomal testing is here: http://www.smgf.org/education/animations/autosomal.jspx

Recently 23andMe started a beta test of a new product, Relative Finder, that will allow persons in their databases to compare autosomal test results and will give you results for genetic cousins with whom you share blocks of genes. The autosomes are all the chromosomes, including the X chromosome, other than the Y chromosome and mtDNA. It appears they will allow people to get in touch with their matches to compare pedigrees. This will allow you to find more genetic cousins than is possible testing only your direct Y DNA or mtDNA line. This also allows females to have matches out side of their mtDNA lines.

The beta test period ends on 30 September 2009. If you are interested in testing, email me and I will put you in touch with some project coordinators who have special codes that give a very low price on the full 23andMe genetic test.

In addition to the Relative Finder program, 23andMe also will provide information on possible health or medical conditions.

This test does not replace the current Y DNA and mtDNA offerings from companies like Family Tree DNA but they supplement that data. FTDNA's database is much larger than that of 23andMe and provides more in-depth testing of Y STR, SNP, and mtDNA full genomes than does 23andMe. Serious genetic genealogists will want to use all of these tests.

Here is the abstract from the American Society of Human Genetics session on Relative Finder's technology: http://www.ashg.org/2009meeting/abstracts/fulltext/f10169.htm

We can look forward to a time when we can check the autosomal markers in databases at SMGF and 23andMe, as well as the Y DNA and mtDNA databases at SMGF and FTDNA.

DNA testing companies I have used

Family Tree Dna:
http://www.ftdna.com/

12, 25, 37, 67 Y DNA markers and advanced 
markers; X chromosome; CODIS; R1a1 Deep Clade; 
mtDNA full genome sequence, H16.

DNA Heritage:
http://www.dnaheritage.com/

43 Marker Y DNA test; R1a1

Trace Genetics:
http://www.tracegenetics.com/

R1a Deep Clade tests.

Coriell Institute:
http://www.coriell.org/

Full genome scan for medical conditions.

23andMe:
http://www.23andme.com/

Scan for medical conditions and genealogical 
information.

Sorenson Molecular Genetics Foundation:
http://www.smgf.org/

37 marker Y DNA; mtDNA; and autosomal DNA testing.