Deeper discoveries with new DNA Match tools | Ancestry®

[AimSmall]

Got my Pro Tools activation of Shared DNA with matches. Will be most useful.

[Mitchell-Atkins]

The one thing I've been doing with the new tool, is identifying which branch of a match's tree that we are connected through by looking at their close matches. In doing this, I did come across something all should be aware of.

If the match hasn't taken the time to set up the tree so the dna tester is the home person you may be looking at the wrong tree/wrong part of the tree.

For example, I had one match that had a nicely filled out tree and she had a close cousin match (363 cM) who also had a filled out tree. Baring and NDE it should be easy to find the shared line, but I couldn't find it.

I then noted that the color of the home person on the tree was pink for female, but on the main page it was blue for male. When I expanded the tree and switched the tree to the match's spouse the shared line appeared. She had herself as home person, but it should have been set to her husband.

In the above scenario, a surname or location search of matches will not be looking at the right tree.

[Stephen1986]

Pro tools are now available on ancestry.co.uk for £4.99 per month.

[Aroon]

Pro Tools are now available in Canada for $7.49 per month.

[Mabrams]

What the heck, I am paying $10 US per month     

7.49 Canadian Dollar =  5.48 US Dollar
4.99 Pound Sterling   =  6.33 US Dollar

That said, with the ProTools, I have made considerable progress on two confusing clusters.   While other clusters remain stubborn.

It would be nice if you could sort your match's match from high to low, as it takes some time to page through them all looking for matches with sufficient cM and/or trees to be of interest.

A couple of bugs.  Saving your notes on certain screens does not happen.  And when I filter matches, I can only see the first page of results (20 matches).

[geebee-1015]

Here's another fun "reveal" based on the ProTools sharing tool.  I've included some who tested not at Ancestry, but at 23andMe; I've marked them as such.  All are descendants of my mother's paternal grandparents.  

The 1st one is 1st cousin to my mother, so a 1st cousin once removed to me:

• JB.  344 cM shared in 16 segments, longest shared segment 33 cM.
  

These two are 2nd cousins, both to each other and to me:

• RGK.  265 cM shared in 14 segments, longest shared segment 34 cM.  In addition to the sharing reported by Ancestry, GEDmatch shows that BK and I also share 41 cM of X-chromosome DNA from our shared great grandmother.
  
• GRK.  261 cM shared in 15 segments, longest shared segment 73 cM.
  
• RB1 (23&Me).  207 cM shared in 11 segments, longest shared segment 37.  BK and RB1 are 1st cousins to each other.    
  

These are all offspring of one or another of my 2nd cousins, so 2nd cousins 1xR to me:

• CK.  196 cM shared in 9 segments, longest shared segment 41 cM.  CK is BK's son.
  
• CBY.  165 cM shared in 10 segments, longest shared segment 38 cM.  CBY is RB1's daughter.
  
• EBJ.  158 cM shared in 10 segments, longest shared segment 34 cM.  EBJ is RB1's daughter, and CBY's sister.
  
• NEK.  148 cM shared in 8 segments, longest shared segment 54 cM.  NEK is GK's daughter.
  
• RB2.  145 cM shared in 9 segments, longest shared segment 41 cM.  RB2 is the daughter of a 1st cousin of BK and RB1.
  
• JPW (23&Me).  127 cM shared in 9 segments, longest shared segment 29 cM.  Also includes a 31 cM segment on the X chromosome, which Ancestry of course does not use and I have subtracted from the total reported by 23&ME.  JPW is the son of a 1st cousin of BK and RB1.  
  
• SKB.  53 cM after Timber; 88 cM unweighted shared in 4 segments, longest shared segment 52 cM.  Also tested at 23&Me, where her sharing is reported as 161 cM.  However, this amount very likely includes a significant segment on the X chromosome.  SKB's father is BK's full brother, and may very well have inherited significant X chromosome DNA from our shared great grandmother.  All but about 20 cM of my entire X chromosome came from this great grandmother. 
  

Look, though, at what Timber does to the sharing between my daughter and three of my full nieces, all of whom are related to the above in the same way -- just one step further than I am.

• JB.  1st cousin 2xR to all four.  KB (my daughter) shares 93 cM in 5 segments, with a longest shared segment of 26 cM.  GL (niece 1) shares 206 cM; SS (niece 2) shares 161 cM; and SJ (niece 3) shares 203 cM.
  
• RGK.  2nd cousin 1xR.  KB shares 116 in 5 segments, with a longest shared segment of 42 cM.  GL shares 136 cM; SS shares 34 cM -- likely after a significant downward "adjustment" by Timber, which Ancestry conceals in their match reporting; SJ shares 143 cM.  
  
• GRK.  2nd cousin 1xR.  KB shares 47 cM after 40 cM is lopped off by Timber; unweighted sharing is 87 cM in 6 segments, with a longest shared segment of 26 cM.  GL shares 189 cM; SS shares 135 cM; SJ shares 111 cM.  If I didn't already know exactly how KB is related to GRK, she would be misled into thinking that he is her 3rd cousin 1xR, or her half 2nd cousin 2xR.  This is the exact opposite of what Ancestry claims as Timber's purpose, but it is a very common result!
  
• RB1 (23&Me).  2nd cousin 1xR.  KB shares 70 cM; GL not tested at 23andMe; SS shares 103 cM; SJ shares 97 cM.
  
• CK.  3rd cousin.   KB shares 45 cM after 16 cM is lopped off by Timber; unweighted sharing is 61 cM, with a longest shared segment of 27 cM.  GL shares 55 cM; SS shares 21 cM; and SJ shares 106 cM.  It is very likely that, like KB, GL and SS have both had downward adjustments by Timber.
  
• CBY.  3rd cousin.  KB shares 25 cM after a 8 cM downward adjustment by Timber.  GL shares 50 cM; SS shares 93 cM; and 101 cM.  GL's true sharing is likely more than this.  GL's mother is my full sister, who shares 130 cM with CBY at 23&Me.  
  
• EBJ.  3rd cousin.  KB shares 52 cM after 15 cM is lopped off by Timber; unweighted sharing is 67 cM in 3 segments, with a longest shared segment of 34 cM.  GL shares 42 cM; SS shares 41 cM; SJ shares 23 cM.  EBJ also has results from 23&Me.  There she shares 131 cM with GL's mother; 151 cM with SS's father -- who is my full brother; and 89 cM with SJ's mother, who is another of my full sisters.
  
• NEK.  3rd cousin.  KB shares 11 cM after 31 cM is lopped off by Timber; unweighted sharing is 42 cM in 2 segments, and the longest shared segment (26 cM) is over twice as much by itself as Timber's made-up number all by itself.
  
• RB2.  3rd cousin.  KB shares 60 cM after 27 cM is lopped off by Timber; unweighted sharing is 87 cM in 3 segments, with a longest shared segment of 41 cM.  GL shares 58 cM; SS shares 28 cM; and SJ shares 41 cM.  But once again, it is likely that all three nieces have had their unweighted sharing adjusted, and as I've showed with my daughter, these adjustments can be very large (47 cM in some cases).
  
• JPW (23&Me).  3rd cousin.  KB shares 96 cM, though this may include some sharing on the X chromosome; GL's mother shares 118 cM; SS shares 80 cM; and SJ shares 126 cM.
  
• SKB.  3rd cousin.  KB shares 35 cM, after 47 cM is lopped off by Timber; unweighted sharing is 85 cM, with a longest shared segment of 52 cM in 3 segments -- 17 cM larger than Timber's reported total all by itself ... without even considering the other 2 segments.  GL shares 85 cM; any sharing of SS and SJ is unreported, which could mean they don't share any DNA with SKB.  However, thanks to Timber it's impossible to know.  My daughter has numerous matches -- some even larger than 20 cM -- which show neither parent as a shared match, even though the matches appear to be real.  This almost certainly mean that Timber has adjusted the parent's sharing to below 8.0 cM, at which point Ancestry simply dropped the match for the connecting parent.
  

I am not saying that Timber never does what it's supposed to do.  I haven't seen any instances in which Timber made any meaningful difference, except in a negative way.  To me that's reason enough that Ancestry should stop bragging on it and stop using it.  Or at the very least, they should use it simply to "caution" customers that a particular amount of sharing might be overreported, rather than to presume that it is.  Their presumption is often wrong.

[Rufus191]

What the heck, I am paying $10 US per month     

7.49 Canadian Dollar =  5.48 US Dollar
4.99 Pound Sterling   =  6.33 US Dollar

That said, with the ProTools, I have made considerable progress on two confusing clusters.   While other clusters remain stubborn.

It would be nice if you could sort your match's match from high to low, as it takes some time to page through them all looking for matches with sufficient cM and/or trees to be of interest.

A couple of bugs.  Saving your notes on certain screens does not happen.  And when I filter matches, I can only see the first page of results (20 matches).

Not sure if someone at Ancestry couldn't do currency conversion, but a few days ago the UK price went up to £7.99! Lucky for those early adopters it looks like they will stay on that rate if they don't cancel of course . Unfortunately that didn't include me, but these tools certainly do look useful as above, very good for identifying clusters, and at least narrows down the part of their tree you link to if you have enough to work from. I should say of course, also, in case it wasn't obvious from the previous posts, but ProTools also gives shared matches (I think) right down to the lowest cM level, which say, if it's only 8cM-8cM might need some caution.

[Riverman]

Searching for ethnicity in common being not included?

[Mabrams]

.Searching for ethnicity in common is not included  at present.  But it will likely be included, so its more a matter of "when"  Including Genetic Communities.
Although I dont think you will need ProTools for this, but you will need a subscription.

Also to be included.  Sorting by the cM of your Shared Matches.  And some level of triangulation. ProTools for these.

A video from RootsTech with an explanation of new features on the horizon.
https://www.youtube.com/watch?v=Q7QAJjYtXLU
at 6:50 is a discussion on Filtering by ethnicity or communites. At 9 min, ProTools is added.

The presentation is not always clear on who gets what.

From Rufus191
"but ProTools also gives shared matches (I think) right down to the lowest cM level, which say, if it's only 8cM-8cM might need some caution."
Actually all the way down to 6 cM, if you have those archived from 2-3 years ago.  So extra caution.   

I am having trouble with the bugs.  It might be that Ancestry will work on these before adding the next feature.... or maybe not.

• Cannot look at filtered matches beyond page 1
  
• Notes often are not saved
  
• Very subtle issue.  Sometimes I use "+" or "-"  to rate trees.  As in "Tree-" for a tree of no value.  The word "Tree" saves sans the "-".  Odd.  Other times the + or - symbol saves correctly.
  

[Rufus191]

From Rufus191
"but ProTools also gives shared matches (I think) right down to the lowest cM level, which say, if it's only 8cM-8cM might need some caution."
Actually all the way down to 6 cM, if you have those archived from 2-3 years ago.  So extra caution.   

I am having trouble with the bugs.  It might be that Ancestry will work on these before adding the next feature.... or maybe not.

• Cannot look at filtered matches beyond page 1
  
• Notes often are not saved
  
• Very subtle issue.  Sometimes I use "+" or "-"  to rate trees.  As in "Tree-" for a tree of no value.  The word "Tree" saves sans the "-".  Odd.  Other times the + or - symbol saves correctly.
  

Yes, that is the case saved 6 and 7 matches are included in the extended shared matches feature. The odds of these being simply coincidental greatly increase at low cMs i.e. you match match A on one line, they match match B and you do, but actually it is from different lines. The SideView paternal/maternal split may help with this but there will still be a high % possibility of coincidental matches at these lower levels.
  Regarding the bug of not getting past page 1 of shared matches - this happens to me, but if I then hit refresh/reload, it works. Regarding notes, when opening matches, I am by default now opening two pages, one in the ProTools shared match page and one in the old style page where it shows the tree and ethnicity, I just edit notes on that old style link to avoid issues.

[Mabrams]

" Regarding the bug of not getting past page 1 of shared matches - this happens to me, but if I then hit refresh/reload, it works. Regarding notes, when opening matches, I am by default now opening two pages, one in the ProTools shared match page and one in the old style page where it shows the tree and ethnicity, I just edit notes on that old style link to avoid issues."

I tried the refresh/reload but no success on the filtered pages of shared matches.   

I am also using a second page in the old style to record and edit notes.   Although I dont know how many notes I lost before I realized they were saving erratically.

[Rufus191]

" Regarding the bug of not getting past page 1 of shared matches - this happens to me, but if I then hit refresh/reload, it works. Regarding notes, when opening matches, I am by default now opening two pages, one in the ProTools shared match page and one in the old style page where it shows the tree and ethnicity, I just edit notes on that old style link to avoid issues."

I tried the refresh/reload but no success on the filtered pages of shared matches.   

I am also using a second page in the old style to record and edit notes.   Although I dont know how many notes I lost before I realized they were saving erratically.

Not sure what you mean by filtered pages? I click the Shared Match Pro link, and I get a page with three columns, the shared match name, the amount the tester connects to that shared match, and the amount the match connects to that shared match. I realised I was incorrect in thinking all limitations had been removed re: cM. The amount in the third column, how much the match links to the shared match is still limited to 20cM, but the amount the tester links to the shared match the limit is removed and now goes down to 8cM. So in fact there are still hidden shared matches which could be useful and quite reliable that you are not seeing i.e. ones the tester match at 20cM, but the match matches at 12-19cM, we still cannot see these. So it still makes sharing of cousin/sibling etc. kits useful to see these hidden shared matches.

[Mabrams]

"Not sure what you mean by filtered pages? "

By "filtered" I was referring to one of two situations.

On the left side, there is a button for Filtered Shared Matches.  Trees, Groups (Dots), Shared DNA and some misc options at the top.  If you pick a filtering criteria that will produce more than 20 results, you will get more than one page.  When I tried to go to the second page, I only saw the first page again. 

About 3 or 4 days ago, the filtering started to work better.  Most of the time I can access page 2 of filtered results.  Yesterday, it worked all day so maybe the hiccups are now a past thing.

The second situation was by using the Search box on the right side.  After searching for a common name, e.g. Smith, I could only see the first page (20 matches).  But that also works better now. 

Saving notes also seems to work reliably now.

Overall, my current experience is that ProTools Shared Matches is bug-free.

Now........if they would just add the Sort to the Shared Matches cM column (right side).  Paging through is very laborious.

[Rufus191]

"Not sure what you mean by filtered pages? "

By "filtered" I was referring to one of two situations.

On the left side, there is a button for Filtered Shared Matches.  Trees, Groups (Dots), Shared DNA and some misc options at the top.  If you pick a filtering criteria that will produce more than 20 results, you will get more than one page.  When I tried to go to the second page, I only saw the first page again. 

About 3 or 4 days ago, the filtering started to work better.  Most of the time I can access page 2 of filtered results.  Yesterday, it worked all day so maybe the hiccups are now a past thing.

The second situation was by using the Search box on the right side.  After searching for a common name, e.g. Smith, I could only see the first page (20 matches).  But that also works better now. 

Saving notes also seems to work reliably now.

Overall, my current experience is that ProTools Shared Matches is bug-free.

Hah, I admit I didn't even notice these options yet . Good to know they might now be fixed...and I will start experimenting further....

[Mabrams]

"Overall, my current experience is that ProTools Shared Matches is bug-free."

Blimey.  I wrote too soon.

The bug on Filtered Shared Matches has reappeared for me. My filtered results only work for the first page or 20 matches.

I made a lot of notes today, and so far they all seemed to have saved correctly.