Visual Phasing: An Example (Part 5 of 5)

Five-Part Series on Visual Phasing:

In “Visual Phasing: An Example (Part 1),” we identified and labeled all of the recombination points in the three siblings, Susan, Brooke, and Felix. Then, in Part 2 of the series, we used the identified and labeled recombination points to assign segments of DNA to the four grandparents (the blue grandparent and grey grandparent pair, and the purple grandparent and green grandparent pair). In Part 3 of the series, we used cousin matching to identify the grandparental source of the chromosomal segments. And finally, in Part 4 of the series, we characterized my paternal chromosome.

After Part 3, we had the following for Brooke, Susan, and Felix:


Now that we have this information, let’s see if we can use that to explore new matches with Brooke, Susan, and Felix.

I’m particularly interested in the DNA from Helen Johnson, my great-grandmother (and grandmother of the three siblings), who was born and adopted in 1889. I have used DNA to help identify Helen Johnson’s biological mother, but I do not yet know who her biological father was.

If I can identify matches to Brooke, Susan, and Felix that are definitively matching through Helen Johnson, they will help me identify Helen’s biological father.

Some of this, of course, I could have done previously. I knew at least one segment that came from Helen Johnson to each of the three siblings, the segment they shared with the match to Helen’s mother’s family. Anyone that matches Brooke, Susan, AND Felix at the region shared with this match (between the dotted red lines), is shared via Helen Johnson’s mother:


But before visual phasing, I was limited to that segment. I had no idea how far that segment extended outward on each of the three siblings. Solely from visual phasing, I’ve learned that Brooke has an entire chromosome from Helen Johnson. The area shown in the dotted black lines is now an area of interest:


If I identify matches from 31 Mb to the end of chromosome 21 that are shared by Brooke and neither Susan nor Felix, I will have identified matches that share ancestry with Helen Johnson.

There are several ways to do this. For example, I can use the Matching Segment Search at GEDmatch to find segments on chromosome 21 from 31 Mb to the end, shared by Brooke but neither Susan nor Felix. I’ll run the Matching Segment Search for each of them, take the results from chromosome 21 (31 Mb to end) and combine them into a spreadsheet.


This will allow me to identify matches shared by Brooke but neither Susan nor Felix. The following is a screenshot of the spreadsheet, showing two segments in rows 8 and 9 that are shared with Brooke but neither Susan nor Felix:


The segment shared with A80XXXX shows nicely in a One-to-One comparison:

Unfortunately these are very small segments at 13.2 and 12.4 cM, but they illustrate the process. And I may find larger segments on other chromosomes.

Now, if A80XXXX and A89XXXX are indeed matches via Helen Johnson’s DNA (and thus match Brooke on a single chromosome), then they MUST share that segment in common with each other. A One-to-One comparison of the two kits at GEDmatch confirms that:


I can contact the owners of kits A80XXXX and A89XXXX, and I can review their trees to find people who were in the right place at the right time.

Although I’ve focused on Helen Johnson’s DNA, I can also use the results for any of the other grandparents. For example, the spreadsheet above shows a very promising group of matches shared by Susan and Brooke but not with Felix. They share a much larger 25 cM segment, which must be shared via their mother and the blue grandparent (whichever of their maternal grandparents this turns out to be).

This is just one way to use the information from visual phasing!

The Future

I envision two major developments with Visual Phasing, although the uses of this output is limited only by our creativity.

1. Visual Phasing Automation.  Visual Phasing will be partially or entirely automated. The data from three siblings could be fed into an algorithm that determines recombination points and fills in the chromosomes. Then, you could feed known matches into the algorithm to have it automatically identify which segments came from which grandparent.

2. Grandparent Reconstruction. The data from Visual Phasing, particularly from people lucky enough to have multiple siblings (4 or more) tested, could be used to recreate incredibly complete and accurate reconstructions of our grandparents. For example, we could extract the Visually Phased segments from one or more of the grandparents to create a grandparent kit. And since we’re using entire portions of chromosomes beyond matching segments, this will allow for greater reconstruction than tools like Lazarus which rely only on matching segments.

Using our visually phased chromosome 21 as an example, from just these 3 siblings we have an ENTIRE copy of chromosome 21 for Helen Johnson and both maternal grandparents (blue grandparent and gray grandparent). We have roughly half of Frank Bettinger’s chromosome 21:


Final Words

I hope you’ve enjoyed this series on Visual Phasing. I’d like to thank Kathy Johnston and Randy Whited for sharing their knowledge with all of us. I’d also like to thank others who have blogged or posted about their experiences with Visual Phasing. I think Visual Phasing is a terrific example of how existing data can give rise to new tools and new information!

Best of luck with your own Visual Phasing project!



2 Responses

  1. Jonathan Long 28 November 2016 / 12:55 pm

    Thanks for sharing this information. Can you add a few notes using 1st cousin results, and how this concept can permit extension back to great-grandparents using combinations of sibling + 1st cousin matches?
    I have two pairs of 1st cousin siblings with test results, and the potential to add a third sibling to one pair, plus another 1st cousin with one set of common great grandparents.
    I assume that it will pay off to add the third sibling to facilitate this. Is the idea that you’d need to reconstruct two individuals who were siblings (based upon their grandchildren’s results) to figure out those individuals’ parents, or three reconstructed siblings to figure out their grandparents? (i.e., three sets of three siblings who are first cousins could potentially allow you to reconstruct their shared great-grandparents?)

  2. Ania 29 November 2016 / 7:15 am

    How does this work if you only have ONE sibling to compare with?

Leave a Reply

Your email address will not be published. Required fields are marked *