Getting Started with Autosomal DNA Part I

So you are thinking about getting an autosomal (atDNA) test, but are not sure where to start. This blog post will walk you through several steps to help get you started.

An atDNA test will result in a list of Matches based on shared DNA. Almost all of these Matches are your cousins – most will be about 5th to 8th cousins, with some who are closer and some who are more distant. The DNA test will give you this list, and a way to contact your Matches; it’s up to you to share information with your Matches and determine your Common Ancestor(s).


  1. Determine your objectives. Write your own or choose from these:

A. ___Find new cousins
B. ___Prove your ancestral lines
C. ___Break down brick walls
D. ___Find biological parent(s) of yourself or some ancestor [see also DNAAdoption*]
E. ___Find out your deep ancestry
F. ___Form working groups of your Matches by Ancestor                                               [Triangulated Groups]
G. ___Determine which ancestor provided each part of your DNA                               [Chromosome Mapping]
H. ___Other ______________________________________
2. What to expect from your atDNA results.
Work. Your results will include a list of Matches – people who match your DNA. In general these Matches will be cousins. Generally very few will be close cousins (1st or 2nd cousins) – the bulk of them will be 5th to 8th cousins, or more. Some will have a Tree or Pedigree or list of surnames posted, but many will not. In general you will need to contact your Matches to determine your Common Ancestor. Adoptees and people with close brick walls will need to compare a lot of information from their Matches to develop common threads, and likely relationships. You need to be involved – your Tree is not magically filled out for you.
Ethnicity is a broad estimate. Your results will also include some estimates of your ethnicity or geographic ancestry. Since you only got part of your ancestor’s DNA, these estimates are generally correct, but not very precise.
Maybe unexpected results. When you take a DNA test and are compared with everyone else who has done the same, there is always the potential for a surprise. You may have an ancestor who is not the biological child of the parents you thought they were.
Genetic Genealogy Standards. This document is highly recommended for more information about DNA testing:
DNA is just a tool – a very power full tool. You use it as part of your genealogy research, not in place of genealogy. A DNA test by itself cannot create your pedigree.

3. Understand the three types of DNA.
Y-DNA used to study an all-male line – the Y-DNA is passed from fathers to sons.
mtDNA used to study an all-female line – the mtDNA is passed from mothers to their children (sons and daughters).
atDNA used to study all of your ancestry – the atDNA is passed from male and female to their children. This test results in Matches from all of your ancestry. The Matches are your cousins – much more on this later. The atDNA does not include any Y-DNA or mtDNA, although Matches could just as easily be cousins from an all-male-line or all-female-line ancestor, as from any other line.

This post is all about atDNA

4. There are two fundamental levels for using atDNA:
Level 1. Genealogy only. At this level you just accept the list of Matches as your cousins, correspond and share with them to determine how you are related. You may often find that you are related several ways. This is plain and simple genealogy. Think of the DNA test as a filter that separates out only people who are related to you.
Level 2. Using the DNA. This level requires some amount of knowledge about how DNA works – how you got it, how your Match got it, what matching means and doesn’t mean, and some amount of jargon. Much more on all of this later. But if it gets too complex, or if you need a breather, just fall back on Level 1, and work with your Matches.

5. Select the company for your test. The base price is $99 at each of the three companies, and they have $10-$20 reductions several times a year. Each company offers a different mix of features. See for a comprehensive and unbiased comparison matrix. Many folks have their favorite companies for different reasons – but this is my blog, so my thoughts include:

A. All three companies display a list of your Matches (people who share some DNA with you and are your cousins, in most cases), and a way to communicate with them. You can and should upload a GEDcom of your Ancestry to each site. They also offer an estimated relationship range for each Match – a range of relatedness (e.g. 3rd to 5th cousins). They show some ethnicity/geographic estimates (none are very precise because you only get part of each ancestor’s DNA). They also give you the ability to download your raw DNA data.

B. Family Tree DNA (FTDNA) – I think is the best all around. Almost all Matches are listed with real names, emails, and all DNA segment data – upfront and easily downloaded to a spreadsheet which you can use or print out. If you have Colonial American ancestry, you’ll probably get more than a thousand Matches. They also store your DNA and offer a range of other DNA tests. If you have elderly relatives, and want to preserve their DNA for future tests, this is the best site. The main drawback is the ability to compare your Matches with each other – this is mostly overcome with their InCommonWith utility. A good site for all the objectives above.

C. AncestryDNA – If you have Colonial American ancestry, you’ll probably get several thousand Matches. Some Matches have good Trees, some have small or no Trees; and some have Private Trees. To communicate with Matches you must use the Ancestry messaging system. Several of Ancestry’s key features include a Hint system which highlights ancestors in your Tree and your Match’s Tree which are the same, based on genealogy. They also provide a Shared Matches feature based on shared DNA; but they don’t provide any DNA segment data, which is essential for objectives B, D, F and G above. A good site for finding cousins, a poor site for working with DNA. Use this site if you don’t want to learn about DNA.

D. 23andMe – has the largest database (over one million customers, but they only list your top 2,000 Matches). There are no emails posted, so you have to use their messaging system to communicate. They have a utility to compare kits to each other, which is a key feature. Their Tree system is the hardest to use. A good site for all of the objectives above.

E. GEDmatch – is a third party site (free, but donations are encouraged) – you can upload your raw DNA data file from any of the above companies to GEDmatch, and compare among Matches who tested at the three companies above. They list the top 1,500 Matches in a One-to-Many utility; and let you compare any two Matches One-to-One. A great suite of other utilities, including Triangulation and several ethnicity/geography programs. I encourage you to upload to and use GEDmatch. You’ll get more Matches and DNA data.

F. My strong recommendation is to test at all 3 companies – each has a different database of potential Matches, and each offers different features. To save some money, you can test at AncestryDNA during a sale, and then upload (copy) your DNA data to FTDNA for $39, upload to GEDmatch (free), and also test at 23andMe to maximize your chances of finding good, close cousins. Balance this plan against your budget and your desire to test other close relatives (also recommended).


6. Develop a robust Tree of your Ancestors. By robust, I mean include as many Ancestors as you can, with place and date information, out to 12 generations or so. This is your “bait” when “fishing” for cousins. The atDNA test tells you a Match shares DNA with you – that they are probably a cousin. You have to compare ancestors with the Match to determine your Common Ancestor. If you both don’t have the Common Ancestor in your Tree, it’s very hard to find it. Most of your DNA Matches will be in the 5th to 8th cousin range, some more distant. You need a Tree that includes as much of your Ancestry as possible back to at least 10th cousin range, wherever possible. Few have actually done the detailed research to “prove” all of their ancestors back that far. My recommendation is to “borrow” from the research and Trees of others to fill out your own Tree as much as possible. I’d go so far as to also include “iffy” Ancestors at the tips of your Tree – ones you may not have researched or proved – these are better than blank spaces in your Tree. The objective here is to identify potential Common Ancestors. Then you and your Match (now a potential cousin) can compare notes to see how much documentation you each have.
Create a GEDcom of your robust Tree and upload it to each site where you’ve tested (FTDNA, AncestryDNA, and/or 23andMe); to GEDmatch; and/or to WorldConnect, WikiTree, FamilySearch, etc.

7. Develop a list of Patriarchs [optional, but very helpful] – make an alphabetical list of your ancestral surnames. Then add the most recognizable Patriarch (or Matriarch, if there is no Patriarch) with years and places. Keep each surname/Patriarch to one line, if possible. Some examples:
   CHILES, Col Walter II 1630-1671; John 1679-1723VA; dau Valentine 1719 Caroline Co, VA
   FISHER, George b c1742 PA; RevWar; descendants to Pendleton; then Harrison/Lewis Co, (W)VA
   HAMM, Stephen b c1737 Amherst Co, VA (on Stovall Creek) early 1700s through RevWar

8. Develop a Standard Message – This is a message you’ll send to all your Matches. It’s good to have a Standard Message (which you can tweak over time). You can just copy and paste it into an email or messaging system. This saves a lot of time. After this initial effort to contact each Match, you’ll want to personalize follow up messages.

Your message should include your real name and email; perhaps a very brief introduction, a link to your Tree of Ancestors, a request that your Match share their Tree with you.

An example (revise to suit your style):
Hi, I’m Jim Bartlett. I’ve been a genealogist since 1974. Most of my ancestry is from Colonial Virginia with one grandparent’s ancestry from Scotland and Germany in the mid-1800s. I’m willing to share my documentation. My goals include validating my ancestral lines and working through brick walls using DNA. My Public Tree: (I can send invite). Please share your Tree (best), pedigree, list of Patriarchs or surnames.

Ask if you have questions – I teach DNA for genealogists; see my atDNA How to Succeed list at: it has some good links at the end; one on Triangulation! Also see my blog:

Hope to hear from you, Jim Bartlett

I have modified my introductory message many time – I’m now on version 23. And you can add in anything, whenever it is appropriate. An example would be something about a particular surname or location you see in a Match’s profile. The “boilerplate” is in your standard message, but you can modify it any time you want.

This blogpost will get you started, and let you order a test with some knowledge of what’s involved. The next blogpost in this series will be:

Getting Started with Autosomal DNA Part II



01A Segment-ology: Getting Started with Autosomal DNA Part I – by Jim Bartlett 20151122


22 thoughts on “Getting Started with Autosomal DNA Part I

  1. Hi Jiim
    I’ve discovered I have a very basic question. It arose when I had assembled 26 different matches all of whom appear to match me on the same segment. And when I run them through the Gedmatch triangulation tool NONE of them triangulares. And it seems none of them match each other.
    My question is this: when you say identical segment are you talking about just the precise spot on the chromosome or the actual genetic matter within it? And when we talk about people having different “dna” are we talking about differences in the type or variety of substance (assuming that dna is at some level a substance) or that the arrangement or “genetic code” is different?
    It’s the only way I can explain to myself how so many people could have contemt of a segment inherited from an ancestor different from each other. There must be an almost infinite way genetic “matter” can differ from human to human.
    Am I on the right track here?


    • Tom, Our DNA is made up of 4 fundamental molecules – they are abbreviated A, C, G and T. A human being has 46 Chromosomes (23 from each parent), which has about 6 billion of these molecules. When science finally mapped our DNA about 20 years ago, they figured out which letter went in each location in a human DNA. Your DNA and my DNA “maps” are over 99% the same. For genealogy purposes, the companies identify about 700,000 specific places which are known to be the most variable – the markers that make us different. This is a sampling technique that keeps the price of a DNA test low. The markers that are tested are called SNPs (pronounced: snips). The problem is that we have two of each Chromosome, and the testing process gets two values for each SNP. The result is a data file with 700,000 specific locations, and two values for each one. There are only 10 possibilities: AA, AC, AG, AT, CC, CG, CT, GG, GT and TT, because the testing process cannot tell which value came from which parent.
      So to your questions:
      Identical segment means the same SNP letter at each location – usually 1,000 SNP locations in a row on one Chromosome #, or more.
      When we talk about different DNA, we mean that the string of 1,000 SNPs is not the same – often much of it is the same, but not the entire string.
      The concept here is that the only way two individuals can get the same long string of identical SNPs is if they both got it from the same Ancestor.
      Everyone has the same long string of DNA molecule locations – the same helix – the same length of molecules in each Chromosome. But like the imperfections in a diamond, some of the molecules are different.
      So when you look at the Chromosomes at GEDmatch – check the color coding – you’ll see most of your DNA does NOT match up with a Match’s DNA, but some segments (actually called Shared DNA Segments) are the same – highlighted with a blue bar.
      So the key to segment Triangulation is you match A on a particular Chromosome segment, and you match B on, roughly, the same (significantly overlapping), segment on the same Chromosome number. Now there could be two outcomes: you match A on your maternal Chromosome and you match B on your paternal Chromosome (the genealogy DNA testing process cannot tell which Chromosome is which) – in this case you’d have a Common Ancestor with A on your mother’s side; and you’d have a Common Ancestor with B on your father’s side – almost certainly these would be very different strings of SNPs and would NOT match each other; the other outcome is that A and B both match on your same Chromosome (again, we cannot tell from the DNA if it’s on your mother’s side or your father’s side) – in this case you got that DNA from one ancestral line (from one Common Ancestor) – the only way that can happen is when all the SNPs are the same in A and B. The way to determine this is by comparing A to B at GEDmatch to see – yes or no. If A and B match THEN you have segment Triangulation.
      Just overlapping on the same area of a chromosome is not sufficient, and oft times leads to errors. By checking A against B, we can resolve it.
      Hope this helps, Jim


      • Jim–Great explanation, and thanks for the in-depth answer. Best summary of dna basics I’ve ever seen.
        I guess I only have one follow up: do the differences between maternal and paternal results in a triangulation explain how I could have zero triangulations among two dozen matches that all match me on the same segment?


      • Tom, What size are your segments? I didn’t get into pile-ups – these are areas, usually of fairly small (under 10cM) shared segments that science has found in many people – they can be very distant. Another possibililty is an area of your DNA where your own SNPs from your two parents cover all the bases – so almost anyone would match you by zig-zagging back and forth. I have one of those areas, which I just set aside, except for over-15cM segments. Jim


    • Tom, Well that is the answer. About half of all the Shared DNA Segments at 7cM are false. How many SNPs? The concept of shared DNA being from a Common Ancestor depends on segments which are long enough to insure that they came from an Ancestor. With Shared DNA Segments below 10cm, and often below 1,000 SNPs, it’s easy for the matching algorithm to pick some SNPs from your maternal side (chromosome) and some from your paternal side – we sometimes say the algorithm zig-zags between the SNPs on your two Chromosomes of the same # to fabricate a “match”. Clearly a zig-zag set of SNPs does NOT come from one of your Ancestors. This is one of key benefits of segment Triangulation – it weeds out these false Shared Segments. It looks exciting at first – to see a bunch of segments all stacked up in one place – but once they don’t pass part 2 of the Triangulation test (matching each other), the house of cards falls. Almost all segments over 15cM are true segments (no zig-zag), so it’s often easier to just use those segments for segment Triangulation. Jim


      • Thanks Jim,
        I went back and re-read your post on whether triangulation always works, and it also helped with my question.


  2. Hi Jim,
    I am new to Genealogy and I love it so far. I have always wanted to do my family tree and found Familysearch very helpful as i was able to find many records of my mother’s maternal side back to 1690. Unfortunately my mothers father was a foundling, he was able to find out who his mother was but never found out who his father was. I convinced my mother to do the DNA test as this might give us some leads to her fathers side. I organised for her to do the FTdna test and last week her results come in. I have to say I am finding it very hard to understand to say the least.

    She has many matches and I am feeling very overwhelmed. For example I found the below persons results with my mother and It said she was 2 generations back from this person. This was the result. Do you read this a person that might be 2 generations back?

    Minimum threshold size to be included in total = 700 SNPs
    Mismatch-bunching Limit = 350 SNPs
    Minimum segment cM to be included in total = 7.0 cM

    Chr Start Location End Location Centimorgans (cM) SNPs
    X 32,213,042 134,487,570 95.0 8,875
    Largest segment = 95.0 cM
    Total of segments > 7 cM = 95.0 cM Actual.


    • Margaret,
      It appears you have also uploaded the FTDNA results to GEDmatch. It also appears you have clicked on the blue “X” for the Match and gotten only the segment for Chromosome X. It’s hard to make a determination based on one Chromosome. Please go back to GEDmatch and click on the blue “A” to get a more comprehensive picture. The “A” link shows matches on the autosomes (chromosomes 1 to 22). A Match with an MRCA (Most Recent Common Ancestor) 2 generations back would be a first cousin, who would typically share almost 1,000cM (not 95cM)


  3. Where do I find the DNAadoption blog/article mentioned above. I find your article easy to read and understand. Trying to figure out a 100+ year old adoption using DNA and need to learn more about using DNA to determine lines to research.


  4. Pingback: Published by Segmentology. org:  Getting Started with Autosomal DNA Part I | GraveSeeker's Diary

  5. Hi Jim,
    My name is Cheryl and I am rather a newbie to DNA. However, I love your articles as they are easy for me to understand and the directions/recommendation to follow are also there.
    I have tested at all three companies, uploaded to Family Tree DNA from the other companies and transfered to GEDmatch. My question is this. Is there any way to tell if relatives come from my mother or father without testing other family members?
    My father is unknown to me, but I once heard he was Irish, he was tall, from Illinois and was in the service in Philadelphia Dec. 1945-Jan. 1946, he was from Illinois, married at some point, opened a restaurant in Flordia, he passed away, and had four children The name, Eddie Murphy, was dropped. I do not know if some or any of his is true. My mother gave me to her sister to raise as she went back with her husbands No half brothers or sisters from my mother are interested in testing.
    I have had cousins from the 3-4th range show up, but they too have holes in their knowledge.
    Since last Dec., I have spent untold hours trying to find anything that might connect me to my father’s family. Perhaps I have siblings there, nieces, nephews.
    Is there anything I can do to help my research along?
    Thank you for any help you may offer.
    Cheryl Defries-Bradley


      • Thank you, Jim, for your input. I was reading additional articles on this site today. Should I trangulate a certain percentage. Over 10? Thank you and Happy Thanksgiving, Cheryl
        Your articles are GREAT.


      • Cheryl

        A 10cM shared segment threshold is a great place to start. Almost all of the segments will be IBD, and you’ll form a lot of TGs. Once you get those pretty well fixed, you will have a good framework and can then drop down to 8cM and see which ones triangulate with the TGs you already have. You’ll gain confidence by starting with 10cM as there are far fewer issues.

        Jim –



      • Thank you, Jim. At 7 cM, I have almost 1000 matches. Surely, one of them must belong to my dad. LOL.
        I am truly enjoying the journey and have learned a tremendous amount of knowledge.
        Once again, thank you for giving me a place to start to hopefully unravel the mystery surrounding my existence.


      • Cheryl, Exactly half of your DNA came from your Dad. About half of your Matches will, too – unless he descended from small families and/or his ancestry immigrated more recently than your Mom’s. Think: half your Matches are on your Dad’s side, unless something indicates otherwise.

        Jim –



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.