Utilizing MyHeritage AutoClusters to Analyze your DNA Matches

AutoClusters are so much fun and can provide tons of information. I’m going to step through how to analyze your cluster matches easily and productively in conjunction with the MyHeritage tools, but first, a little light housekeeping.

First, please note that this article was presented as a webinar for MyHeritage as part of the MyHeritageDNA Facebook LIVE series. You can watch it anytime, free, at the permanent link, here, courtesy of MyHeritage and me. However, everyone learns differently, and some people do better with written instructions. You can follow the step-by-step instructions in this article.

Second, AutoClusters are a built-in advanced DNA tool at MyHeritage for customers who either:

I would encourage the subscription because many of the MyHeritage tools function best with a large tree. While MyHeritage does offer free trees of up to 250 people, to take full advantage of your DNA test plus tools, you’ll want a larger tree. Subscription features and pricing can be found here and you can try a free trial subscription here.

Third, if you’d like to transfer your DNA file from another vendor, I wrote step-by-step instructions, here.

Fourth, MyHeritage is having a $49 Halloween DNA sale, here, with free shipping if you purchase 2 kits.

And last, Genetic Affairs, the author of AutoClusters, provides additional functionality on their own website for use with FamilyTreeDNA and 23andMe. Customers at Genetic Affairs cannot access MyHeritage data from the Genetic Affairs website since MyHeritage contracts with Genetic Affairs to provide AutoClusters directly to MyHeritage customers at no additional charge. I only mention this because the functionality described in this article and in the companion webinar discusses the functionality by using a combination of AutoClusters and the unique tools available only at MyHeritage.

Ok, housekeeping complete – on to AutoClusters!

Get yourself a cup of coffee or tea. We’re taking a deep dive here, beginning to end, but keep in mind that you don’t have to do everything that’s possible initially, or ever. It’s OK to take baby steps. Just know that AutoClusters can be a superpower to breaking down brick walls. Not only that, AutoClusters are simply FUN!

Let’s start with a basic question.

What is an AutoCluster and Why Do I Care?

An AutoCluster is an artful bouquet of hints, arranged by family group in a puzzle format.

AutoCluster technology, a form of genetic networks, is a way to display your matches who match you and who also match each other in a meaningful, colored-coded group. Each group, or cluster, shares a common ancestral line, somehow. That “how” discovery, or better stated, “which ancestor” discovery is up to you – but clusters provide huge hints!

We’re genealogists, right – we live for hints. Let’s take a look at how this works.

I would suggest reading through this article the first time, then working through the steps as you read it a second time with your own AutoCluster. Don’t worry, I’ll show you how to request one.

This example of my own AutoCluster report, which I’ll be using throughout this article, shows three different clusters.

Everyone within a cluster matches you, but not everyone matches each other. Each cluster is represented by colored cells, each of which represent the intersection of two people who match each other. In the third yellow cluster, everyone matches each other except for two people who don’t match each other.

Grey cells fall into both of the two clusters they are between. For example, the grey cells to the right of the red cluster in the red box match people in both the first red and second tan cluster.

What this means is that once you’ve identified the genesis of each cluster, you know that people who are grey members of both clusters descend from both lines which could represent the two people in an ancestor couple. In my tree, my maternal great-grandfather Joseph Bolton married Margaret Claxton/Clarkson, and I expect the grey people descend from this couple or from both lines individually. One way or another, they match people from both clusters.

The grey people are an additional hint – so don’t neglect them. In fact, some of these grey squares can be even more important that people within clusters because they span two clusters.

Ok, so how do I generate an AutoCluster at MyHeritage?

Requesting an AutoCluster

You’ll find the AutoCluster featured under the DNA menu, under DNA Tools.

Click “Explore.”

If you manage multiple kits, be sure to select the right kit for the right person.

In my case, I have a transfer kit, then I tested at MyHeritage for the health product, so I have two kits. A MyHeritage kit shows with the MH prefix, while a transfer kit shows a different prefix.

The matches and AutoClusters are slightly different between the two kits because the tests are run on different DNA chips.

click to enlarge

After selecting the correct kit, just click on the purple “Generate” button. Note that if your parents have tested, generating an AutoCluster for one or both of them will help you immensely with your own AutoCluster. If both of your parents have tested, you may want to work only with their AutoCluster reports, and not your own. They will have people in their clusters that you don’t because you didn’t inherit that particular piece of DNA from your parents.

Next, you’ll see a message informing you that your AutoCluster is being generated and will be sent to the email registered to your account.

Queue up Jeopardy countdown thinking music

Just a few minutes later, my AutoCluster arrived in my email box. (Note – check your spam folder.)

If you request multiple AutoClusters for different tests or accounts at the same time, take care not to mix them up. Voice of experience here…

You’ll receive 3 items in zip file. I save my files to my computer.

  • Readme file
  • HTML (with the colored circle)
  • Spreadsheet which is a different format of the html file

I don’t know how well the HTML file and the spreadsheet will display on non-computer devices, although I know the HTML file does display on an iPad. I generally work from my computer.

The HTML File

Just click on the HTML file to display your AutoClusters. You’ll get to enjoy seeing them “flying into place,” assembling into clusters. I told you these were fun!

You can play around a bit with options, but “cluster” is the default view and the only one we’re covering in this article.

Each colored cluster is a group of interrelated matches.

I have a total of 18 clusters.

Scroll towards the bottom to view the parameters used to generate the clusters.

click to enlarge

These parameters are not adjustable and have been optimized by MyHeritage to perform well for all customers, including testers with significant endogamy, such as people with Jewish heritage. At the system-generated threshold, I have 100 qualifying matches. Note that the system optimizes the thresholds individually for each person, and your thresholds might be slightly different than mine.

  • Min threshold 40 cM (often this level of match is in the 5C or more distant range)
  • Max threshold 350 cM (closer than 350 would probably be 1C or closer)
  • Shared DNA match minimum threshold 15 cM (overlap of matching DNA)

You’re probably wondering – where are the highest matches such as parents, siblings, uncles, aunts, etc.?

Close family members would be in many clusters. Placing one person into more than two clusters is simply not technically possible due to the constraints of a two-dimensional grid medium, so close family matches are excluded from clusters as to not be confusing. You can still use close family members in shared matching. In fact, they are extremely useful and we will discuss that shortly.

Fly your cursor over the cluster to view the cluster members and their match status to each other. In the grid, each person who matches another has a colored cell. In this example, my cursor is pointing to the cell where “cro” matches Bonnie. Names are obscured for privacy.

Scroll on down below the cluster box to view additional information about each member of the cluster. Many people don’t realize there’s more because they are excited about viewing their clusters and miss this important information about the cluster members beneath the grid.

click to enlarge

Your notes are critically important and you can search by notes. When I identify how someone is related to me, or even clues, I record that information in the notes. I SHOULD have recorded “TOFR” for the matches who have Theories of Family Relativity, and I have gone back and done that now. We’ll talk about TOFRs in a minute.

You may be able to identify the common line or ancestral couple based on the matches alone. Note that these matches may not all be from the same generation. For example, I have some matches in this group who descend from various Claxton ancestors, spanning at least 4 generations. That commonality is how I know the cluster is “Claxton/Clarkson” and not from one of their wives – at least to the most distant generation where I’m stuck.

Matches can span many generations in a “line” and probably involve multiple DNA segments, especially in larger clusters.

Click on “Tree” to view the tree of your match.

Click on “Name” to review their DNA match with you.

Note that your match may match you on more than one line and possibly on both parents’ sides. Inclusion in this cluster simply tells you they match on this line and does not eliminate any other lines.

Now, let’s begin our cluster analysis and drill down.

Select the Best Match

I always begin my analysis with what I think is the “best” match in a cluster.

  • Best could be the largest tree.
  • Best could be the largest match.
  • Best could be the largest number of ICW (in common with) cluster matches.
  • Best is any match with a TOFR (Theory of Family Relativity)

I make notes for all TOFR matches, after verifying, of course, indicating the common ancestors. I also note “TOFR” so I know, when looking at clusters, why I assigned that specific ancestor. When you have a TOFR, MyHeritage has already done the heavy lifting for you.

I note matches’ inclusion in a cluster to remind me to check those clustered matches first. When a match is in a cluster, AutoCluster has done the heavy lifting for you.

The key to success is to utilize multiple tools, together.

Like what?

The Success Triumvirate

Successfully identifying clusters, ancestors and how each person matches you is accomplished through a combination of three primary tools. I call this the “Success Triumvirate” because the three are quite interwoven.

We are going to use all three of these tools, together, so let’s talk about them individually briefly.

Theories of Family Relativity (TOFR)

TOFRs are super hints – theories about which common ancestors your matches share with you.

I wrote about Theories of Family Relativity complete with step-by-step instructions:

TOFRs connect you to your DNA matches by identifying a potential ancestor through a succession of trees and documents from different sources. You can do a number of things to help TOFRs, (and yourself), along.

  • TOFR formation requires a tree, so create one at MyHeritage, using their free TreeBuilder on your computer, or upload a tree that you’ve already created elsewhere.
  • TOFR does best if you complete the tree through grandchildren of each ancestor, at least, if possible, for each generation. Think of each person as a hand reaching out to latch on to the same person in another person’s tree. The more hands, the better your odds of success.
  • Include birth/death date and location, or as much as you know.
  • Accept Smart Matches where appropriate.
  • Make notes. Notes keep you from retracing your own steps.

click to enlarge

A TOFR connection as offered by MyHeritage may not be exactly accurate, but the common ancestor may be accurately identified anyway. For example, in the above TOFR, Margaret Claxton did not marry William Luke Monday, her sister did. The TOFR isn’t exactly correct, but the common ancestors are easily identified. I can take it from this point – no problem.

Always check to see if multiple TOFR paths exist because important hints may be hidden in those links. Think of yourself as a sleuth😊

Let’s take a look at one cousin in this Claxton cluster, Bonnie. What can we learn, and how? Let’s review Bonnie’s DNA match to me.

Reviewing Bonnie’s DNA Match

Clicking on “Review DNA Match” with Bonnie shows me a host of information divided into sections, beginning with a TOFR.

Bonnie Has a TOFR – Hot Diggity!

The first thing we see is that Bonnie does have a TOFR with the tester (me), so we can identify a potential common ancestor.

click to enlarge

Not only that, but Bonnie has a fairly robust tree of 4043 people, so she must be interested in genealogy at some level.

click to enlarge

Not only that, but there are two separate potential “paths” that connect me and Bonnie at a potential common ancestor. One may be more accurate than the other. Be sure to check all paths.

I can click on the little green dots that bridge trees by connecting what the system believes to be the same ancestor to view and evaluate that information.

click to enlarge

Clicking on this green dot would display the match criteria from both trees.

In this case, the weighted match was 76%. The information for Margaret herself was mostly the same, but her husband(s) and children were different due to the inaccuracy of showing her married to her brother-in-law.

Evaluate all TOFRs, links, trees and hints for accuracy. They aren’t gospel.

Another great source of hints is Smart Matches. You may, and probably will, have Smart Matches with people’s trees who are not DNA matches to you. Smart Matches are not necessarily connected to DNA matches specifically, but they do help TOFR form accurately.

Bonnie Has Smart Matches!

MyHeritage generates Smart Matches WITHOUT factoring in genetic matching. Smart Matches occur when enough common factors exist between a person in your tree and a person in another tree whether you are a DNA match with that person or not.

If you have Smart Matches with a DNA match, they will be listed when you review your DNA match with that person.

click to enlarge

To determine whether or not this Smart Match could be relevant to your DNA autocluster, be sure to notice whether this is a direct ancestor of both people. To be relevant to DNA, the Smart Match must be for a direct ancestor or at least lead to a direct ancestor.

Next, click “Review Smart Match.”

click to enlarge

The first thing you’re going to see is “Confirm Match,” and as a genealogist, that stopped me dead in my tracks.

That’s skull-and-crossbones frightening. I don’t know what “confirm match” means or does? Does it mean that all of their information will automatically be copied to my tree if I click that button? I certainly DON’T WANT THAT!!!

I may not want the “Improved Info” either. That information may not actually be improved. What do I do?

For a long time, I did nothing because I didn’t want to mess something up – but doing nothing isn’t the right answer either – because confirming Smart Matches helps TOFRs for everyone.

I wish MyHeritage provided a bit more information here, because “Confirm Match” doesn’t import any information into your tree automatically. You have the opportunity to review everything first.

There are two questions at this point you need to ask and answer independently:

  1. Is this the same person?
  2. If so, do I want any of this data to be imported to my tree?

If it IS the same person, go ahead and confirm – you’ll get to review each new or “improved” item at that point.

If it’s NOT the same person, scroll to the bottom of the page and reject the match.

In this example, Nicholas Speak is the same person, so I’ve clicked on “Confirm Match” which then allows me to review each piece of information that is different, individually. If I want to import that information into my tree, I click on the little arrow to bring the information into my tree, replacing mine. If I do nothing, no information is copied to my tree. It’s that simple. If I make a mistake, I can always edit my own information.

click to enlarge

Bonnie Has Shared Ancestral Surnames!

Another hint for DNA matches is “Shared Ancestral Surnames.” If you can’t figure out how you are related, take a look at these. Of course, Smith is extremely common, but groups of shared surnames are a huge hint, especially if you also have shared locations.

click to enlarge

You may discover more than one line that connects you to this tester – which sometimes makes things a little more complicated! That’s when location can become a life-saver.

Bonnie Has Shared Ancestral Places!

Shared ancestral places can be very useful, even if you can’t identify common surnames, especially in cases where surnames may not be useful. Unknown parent events and adoptions have always occurred, and a specific location may go a long way in terms of identifying the ancestors of both parties that may be related.

Purple pins with numbers mean you BOTH have ancestors from that location. Bonnie and I share 65 ancestors from one place. I definitely need to evaluate that location!

click to enlarge

Clicking on Tennessee shows the pins in that location. Clicking on a specific pin displays the ancestors from that location.

Note that the purple “65” pin location revealed this common ancestor whose surname is spelled differently in our trees. This surname transitioned back and forth, so there I no “right” or “wrong” way to spell it. However, a different spelling may keep the person from being recognized as the same individual by computer software.

Now, let’s review Bonnie’s shared DNA match information.

Bonnie’s Shared DNA Matches

We know that each of the people in the first cluster match the tester, me, and all but 3 (yellow stars) of the people who match me in the first cluster also match Bonnie

However, don’t think for one minute that there are only 8 people who match me and Bonnie both. There are only 8 who match us both AND are included in the cluster. These are judged to be out “best” common matches.

Looking at my DNA match with Bonnie, I see that there are 162 total shared matches.

The balance, other than the 8 in the cluster, did not meet all of the match threshold ranges to be included in the cluster. In other words, shared matches not in the cluster were either less than 40 cM or more than 350 cM, or the shared piece of the matching segment was less than 15 cM. In other words, the matches in the cluster are the strongest shared matches, other than close relatives, but they certainly aren’t the only shared matches.

I match Bonnie on two segments, one on chromosome 13 and one on chromosome 16.

Just because someone matches me and Bonnie, both, doesn’t necessarily mean the match is on the same segment. For example, they could match me on chromosome 10 and Bonnie on chromosome 1, while Bonnie and I match each other on chromosomes 13 and 16.

However, there’s certainly a good chance that someone matches us both on the same segment(s).

Reviewing the cluster matches between me and Bonnie, we discover the following information regarding these two specific segments on chromosome 13 and 16, only.

Shared Match with Bonnie Triangulation Chromosome & Location
Sharon Yes Chr 16 only
Renee Yes Chr 16 only
Wilma Yes Chr 16 only
John Yes Chr 16 only
Celeste Yes Chr 16 only
Shirley No Neither
Carolyn Yes Chr 16 only
Ray No Neither

Six people match me and Bonnie both on chromosome 16, none match me and Bonnie both on chromosome 13, so that means that both Shirley and Ray match both of us on a completely different chromosome segment.

Now, of course, the question becomes if those 6 people match Bonnie and me on the same or at least an overlapping portion of chromosome 16.

Triangulation

Triangulation, which I wrote about here, occurs when the tester matches two or more people on the same reasonably sized segment of DNA, and they also match each other on that same segment. The “matching each other” part is important, because it verifies the match is from the same side, Mom or Dad, and from a common ancestor, not identical by chance (IBC).

I wrote about identical by chance here, but in essence, IBC means that a piece of your Mom’s DNA and a piece of your Dad’s DNA accidentally combined in you to look like a match with someone else, but it’s a false positive. You do technically “match” that other person, but it’s because of chance recombination, not because you share DNA from a common ancestor on one side of your family or the other.

The matching to other known family members on that segment is the clue to eliminating IBC matches from comparisons. Each of your valid matches will match one of your parents, or the other. If your match doesn’t also match one or the other parent, it’s not a valid match.

This is known as parental phasing and is why it’s extremely important to have both or one of your parents test, if possible.

If the tester’s parents have tested, each of your cluster matches will match to one parent or the other in addition to the people in the cluster.

Bonnie Has Triangulated Matches!

At MyHeritage, when you review shared matches, you can see if your match triangulates with you by the presence of a little purple triangulate icon.

click to enlarge

Looking at my shared match list with Bonnie, I see Wilma has a purple icon, indicating triangulation between Wilma, Bonnie and me. Woohooo!

Clicking on the purple triangulate icon shows me the common triangulated segment(s).

In this case, Bonnie, Wilma and I only triangulate on one segment, on chromosome 16. Do the other cluster members also triangulate with Bonnie, Wilma and me on this segment? The ones who have a triangulation icon should since I’ve already determined that they only match me on chromosome 16 in common with Bonnie. Let’s see.

click to enlarge

I added the other people in the match cluster to see who else triangulates on any portion of chromosome 16. Just type the names from the cluster into the DNA match name box below the profile cards in the chromosome browser to add each person to the view.

Only the triangulated portion for all people compared is bracketed. That’s so important to remember. In the example above, all people match me and each other on the bracketed portion of chromosome 16.

In this example, two of the people compared do NOT triangulate on this segment, so no bracket is drawn. This might lead you to think that the three people whose DNA matches the tester on the same segment don’t also match each other – but you can’t assume.

If you remove the two people not matching on the segment from the chromosome browser, the other three now show the triangulation bracket.

Triangulated segments provide evidence that a specific segment descends from a common ancestor. The challenge, of course, is to identify the ancestors who contributed that segment generationally through time.

I wrote about triangulation at MyHeritage in the article Triangulation in Action at MyHeritage.

Downloads

You can only compare a maximum of 7 people at a time in the chromosome browser, but you can download your entire match list in a spreadsheet and work from there. I do that often.

There are three different downloads that provide different information and serve a different purpose.

Chromosome Browser Match Download

Scroll down to the bottom of the chromosome browser page to download the matching segments (to you) of the people shown on the browser at that time.

You can download the segments for the current matches showing in the chromosome browser by clicking on advanced options on that page.

Click on “Download shared DNA info.”

This download will happen immediately to your system. I use this technique when chromosome painting matches identified to a particular ancestor at DNAPainter. I also note for each match when I’ve painted their matching segments so I don’t waste time doing it twice.

The second and third download options are found on your DNA Match page.

Export Full Match List and Shared DNA Segments

By moving to your main DNA match page, you can download:

  • Your DNA match list which downloads information about each of your matches
  • Your matching DNA segments for all matches

By clicking on the three dots, you will see the two download/export options. Those two files hold different information.

The “entire DNA matches list” provides information ABOUT your matches, such as:

  • Name
  • Age
  • Country
  • Contact link
  • DNA manager
  • Status (new)
  • Estimated relationship
  • Total cMs
  • Percent shared DNA
  • Number of shared segments
  • Largest segment
  • Link to review DNA match
  • Has tree (yes/no)
  • Number of people in tree
  • Tree manager
  • Contract tree link
  • Number of smart matches
  • Shared ancestral surnames
  • All ancestral surnames
  • Notes

This is important, and I use this file a lot because it provides all of the information in one place and I don’t have to click on each match to evaluate. Plus, I can search and sort to my heart’s content.

Option two, the entire “shared segment DNA info” match list will show all matches, including maternal, paternal and IBC. It’s up to you to figure out which are which, but we have lots of tools and hints.

Your shared segment spreadsheet provides information about the shared DNA, only.

Let’s start by looking at Bonnie again.

Bonnie and Chromosome 16 on the Spreadsheet

Here are my two segment matches with Bonnie in the spreadsheet.

The MyHeritage tools, combined, provide you with the ability to sort your matches meaningfully into genealogically relevant clusters and identify ancestors. I’m going to utilize that information with the downloaded spreadsheet segment information.

Let’s take a look at that matching segment with Bonnie on chromosome 16.

In the shared DNA segment spreadsheet, I filtered for chromosome 16, sorted in lowest to highest order (end location, then start) and looked for matches that fall between these two locations.

In reference to the match with Bonnie, look for any match between 79914629 and 87713399.

I am showing only a partial list below. The actual number of matches to be on this segment of chromosome 16 is about three times as large as this graphic.

After downloading the spreadsheet, I added a Triangulation Group column and a comments column, at right.

click to enlarge

I’ve colored the cluster members yellow who match on chromosome 16 to Bonnie AND me in the cluster.

People who match me on chromosome 16 and are NOT in the cluster fall into one of the following categories:

  • Also match to me and Bonnie, but outside of the cluster threshold. You can see that there are a lot of matches below 40 cM, which immediately eliminates them from the cluster.
  • Match me and Bonnie, but on an overlapping piece of DNA not large enough to be included in a cluster – in other words, the overlap of the three people is less than 15 cM..
  • Match to me, but not Bonnie which means that either they are a match from the other parent’s side, or identical by chance.

Discerning which category each match falls into requires looking at each match and evaluating individually.

You can look at each spreadsheet row, individually, below, if you wish, but what I’d like for you to do is to focus on the groups that I created as I analyzed each match on the segment of chromosome 16 where I match Bonnie.

click to enlarge

  • Green row is Bonnie, our baseline person whose match is why I’m analyzing this particular segment.
  • Bright yellow shows the 6 AutoCluster triangulated chromosome 16 cluster members.
  • Lighter yellow rows are also matches and triangulations on the same segment with me and Bonnie, but not included in the AutoCluster
  • Pink indicates matches on Mom’s side on this same chromosome segment. Mom is in the database, so this is easy to discern.
  • Grey is IBC (darker) or likely IBC (lighter) meaning they don’t match either parent’s side entirely.
  • Bright red is a breakthrough!

You’ll notice that the “best” matches, meaning the ones in the cluster, are clustered together on the spreadsheet too.

The second group of matches, below, begins to have more IBC and matches to Mom’s side. A third group, which I’m not including here, is almost entirely Mom’s side.

click to enlarge

When I finished analyzing the matches on this segment of chromosome 16 from the AutoCluster, I had:

  • Bonnie (green) + 6 Claxton matches (bright yellow) reflecting the first cluster triangulation with Bonnie and me
  • 93 total people that matched me on some portion of chromosome 16 that I match in common with Bonnie. However, on this spreadsheet, matches to me on this segment include some matches who will not match Bonnie.
  • People not matching me and Bonnie both on this segment will include both matches on Mom’s side (pink), and IBC (grey).
  • A breakthrough (bright red) identifying this segment as Claxton, as opposed to Sarah Cook’s, James Claxton’s wife, which means that I can focus on other people with trees with common ancestors who match on that segment on Dad’s side. Someplace in those trees is the information that will someday identify James Claxton’s parents/ancestors.
  • Identified 28 (light yellow) paternal matches through this segment assigned to Claxton that match me and Bonnie, both.
  • Identified 30 (pink) Mom segments, some of which are Acadian and some of which are German. On Mom’s side, two different portions of chromosome 16 recombined from two different ancestors and I can tell where that dividing line occurs by using visual phasing and triangulation at DNAPainter.
  • Identified 26 (grey) IBC segments which are false positive (or likely false positive) matches and should be disregarded.
  • Made notes on each of those matches at MyHeritage.
  • Painted each valid segment at DNAPainter.

About That BreakThrough…

Why is this breakthrough important, and what does it tell me?

Bonnie is descended from the same Claxton line as I am, meaning she is a proven descendant of James Lee Claxton born about 1775 and his wife, Sarah Cook through their son, Fairwick/Fairwix Claxton. We don’t know where James Claxton was born, but likely in either VA or NC. He first appeared on the tax list in Russell County, VA, with no other Claxton males, not long before he married Sarah in 1799.

Bonnie and I match Jim on that same segment.

Jim’s ancestor was Solomon Claxton, born in 1801 in NC. In other words, Jim does NOT share James Claxton as a common ancestor. This means that Jim and Bonnie and I share DNA from a common Claxton ancestor. That segment of chromosome 16 cannot be from the Cook side, because Jim does not descend from the James Claxton/Sarah Cook line.

Therefore, other people who triangulate on that segment, who don’t show trees with Claxton ancestors, and have matching trees to each other will one day hold the key to our common ancestors who contributed that segment to all of us on chromosome 16.

That means I need to take the time to evaluate every one of their trees looking for their common ancestors with each other. It’s likely that common ancestor could be mine as well, or lead to mine.

Just One!!!

Remember, all of the discoveries above were made from analyzing just one chromosome segment match from the Bonnie row in the first AutoCluster. Just one!

Autoclusters intentionally only utilize your “best” non-close family member matches. This allows you to see the genetic relationships between multiple people, even without trees.

You then use the trees, TOFR, surnames, locations, Smart Matches, shared matches, triangulation, and previous research to identify the ancestral connection.

Just scanning this AutoCluster report, I can immediately discern that people share matches between groups of clusters. For example, clusters 1, 2, and 4 share members – for starters. That tells me that these clusters are related to each other. In fact, that’s exactly correct as shown after analysis when I was able to assign each cluster to either an ancestor or ancestral couple.

I discovered a HUGE amount of information researching just one common segment with one match, including a breakthrough which may, one day, if not today, lead to the identification of James Claxton’s parents.

Just think how much more there is left to discover! I need to review the match to Bonnie on chromosome 13 and the other 99 people in my AutoCluster, utilizing the same tools and techniques.

I can hardly wait to get started!

Clusters are Genetic Super-Powers

Clusters are your super-power matches. Take full advantage of them.

  • Every cluster tells a story.
  • If you can identify the common ancestors with one or two people, and it’s the same line, you’ve probably identified the genetic “cluster.”
  • Every match tells a story.
  • You may triangulate on multiple segments with different people.
  • Every individual segment tells a story
  • Each segment stands alone, meaning one segment can descend from the mother of the couple, and another segment from the father. Don’t assume that each shared segment descends from the same ancestor.
  • Don’t assume that if you match one person on two segments, that they both necessarily descend from the same line or couple. It’s possible that you are related on another, known or unknown, line.
  • Every segment match has an individual genealogical history that can lead to different ancestors, meaning that the genetic line is the same, but the ancestors may be different. You may match one person who descends from the son of another match, for example.
  • Each triangulated segment descended from common ancestors who contributed that segment to all triangulation group members.
  • The history of brick walls is held in unidentified matches to segments.

An example is worth 1000 words.

Walking Back In Time

Based on multiple triangulated matches to various people, the triangulated segment on chromosome 16 belongs to the following ancestors:

Generation Ancestor Via Match to…
1 Dad Assigned to Dad’s side via triangulated matches to known relatives
2 Ollie Bolton Culley, Stacey
3 Margaret Clarkson Fred, John
4 Samuel Claxton Wilma
5 Fairwick Claxton Joy, Eugene, Billy, I.B., Bonnie
6 James Claxton, Sarah Cook Brent, Delilah
7 Unknown Claxton parents Jim (NC), Kelsey (TN)

As you can see, based on the genealogy of my matches, I’ve walked the segment on chromosome 16 back in time 7 generations.

How do I get to generation 8?

Clusters are Genetic Super-Powers

Now I need to search the trees of matches on this same segment, but without identified common ancestors to me, looking for common lineages in their trees with each other.

This Claxton segment descended from some unknown ancestor(s) upstream of James Claxton. The key to the identity of those ancestors is held in their DNA segments and matches.

What I’m looking for are common ancestors of those chromosome 16 matches to each other. For example, if James Claxton’s father was named John Claxton and his mother was Jane Doe, finding several people with trees connecting to the Doe family would be especially relevant. Those are the more deeply hidden clues.

I need to do the exact same thing, following the same process, with each segment of every cluster match!

The solution to brick walls is held in unidentified matches to triangulated segments which point the way – like invisible “this way” arrows through that door from our ancestors.

AutoClusters are the genetic superpower!

_____________________________________________________________

Disclosure

I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.

Thank you so much.

DNA Purchases and Free Transfers

Genealogy Products and Services

Genealogy Research

17 thoughts on “Utilizing MyHeritage AutoClusters to Analyze your DNA Matches

  1. Roberta, I’m sorry to ask a question that is not directly connected to the topic of this post. However, I see that, like me, you have autosomal results at MyHeritage both for an FTDNA Family Finder test that you transferred to MyHeritage and a test you did at MyHeritage to get the health information. Do you see differences like I do in the My Heritage ancestry estimates for the transferred test when compared to the MyHeritage test? Here’s what I see:

    FTDNA Family Finder Kit (test taken in 2016)
    Europe 100.0%
    North and West Europe 85.3%
    Scandinavian 44.2%
    English 41.1%
    South Europe 14.7%
    Iberian 14.7%

    MyHeritage Kit (test taken in 2019)
    Europe 100.0%
    North and West Europe 93.5%
    English 59.3%
    Scandinavian 34.2%
    South Europe 6.5%
    Iberian 6.5%

    Some of these differences appear significant to me. I think I have an idea of what might explain the differences but I’d be interested in your ideas if you have the time and interest to respond.

    • I haven’t looked, but I would presume that they are. The chips are different, and I don’t know if ethnicity is determined before or after imputation, but my presumption would be before.

      • Just what I was thinking. Glad to hear we’re on the same page. Matches are different as well. I have 11,583 matches with my FTDNA kit and “only” 8,527 with MH kit. I’m guessing that I should probably focus on the MH kit for AutoClusters.

  2. I have read every line of this post and hope I can use the information to unravel my DNA connections. I am currently using DNA Painter and Genetic Affairs also. I have tested at the major testing companies. The deeper I delve into my genealogy, the more confused I am. So many matches that don’t make sense, the lack of any matches for known documented ancestors who apparently aren’t my ancestors now. A very tight cluster to mostly Finland which includes 2 Smiths from GB but no tree connections and no Smiths in the Finland matches. Just one question today, is there any relevance to the order of the clusters or is it just determined by the amount of matches. This mystery Finland cluster is cluster No. 1 and has no gray boxes. I really enjoy all your posts and the great information you provide. Kathy

  3. Roberta mentioned a bit about MyHeritage optimizing the Autocluster tool for “perform well for all customers, including testers with significant endogamy”. I thought I’d mention my experience. My family could be a textbook case on when autoclusters work… and when they don’t.
    I transferred my parents’ Family Tree DNA results to MyHeritage; perhaps testing there would be different, but I think it would be the same.
    My mother’s results look a lot like the charts displayed above; a number of small different-colored clusters of varying sizes, with grey boxes scattered around. I’ve found several leads and some new cousins, and the tool is worthwhile for her side. She is descended from several different European populations, for the record.
    My father, on the other hand, is entirely descended from one endogamous group(low-German Mennonite), and so far, I’ve found no instances of his ancestors marrying outside that group. When I viewed his autocluster chart, there was one red cluster filling the entire chart. No other clusters, no grey boxes; everybody is related to everybody. Some people don’t match one another, but I’d estimate that the chart is composed of at least 80% red squares.The person who matches the least number of other relatives still matches more than 30 people on the chart.
    Like I’ve said, I’ve found autoclustering to be a very useful tool, and it has been great for my mother’s side of the family, but from my experience, it definitely struggles with endogamous populations. It isn’t any fault with the tool, and there is probably no way it could work well with a family like my father’s ancestors, but it is something to be aware of if you’re descended from a very close-knit group as well.

  4. Thank you for sharing and all the effort into your blog. Now I will have some reading material to keep me occupied for a while.

  5. Pingback: Genetic Genealogy at 20 Years: Where Have We Been, Where Are We Going and What’s Important? | DNAeXplained – Genetic Genealogy

  6. Pingback: Genetic Affairs Instructions and Resources | DNAeXplained – Genetic Genealogy

  7. i didn’t understand how to do it… I have autoclusters for me and both my parents. but this is very advanced level. or maybe it’s just for people with thousands of people in their trees.

Leave a Reply