Author Topic: DNA what is the bare minimum in % for matches cms or segments ?  (Read 972 times)

Offline Shay123

  • RootsChat Extra
  • **
  • Posts: 12
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: DNA what is the bare minimum in % for matches cms or segments ?
« Reply #9 on: Sunday 05 July 20 16:05 BST (UK) »
Hello,
When is a science not a science? When it is biological! :)
GedMatch recommends over 7cM if you are using that criteria alone. The smaller you go the easier it is to get a match with anyone. I think I read elsewhere that 3cM gets that kind of result.
However, your percentage match may also matter. By trial and error initially, I realised that chasing links of 0.3% and less was a similar exercise. Yes, you can make connections but you are going way back and unless you have a really good, expansive tree to use, it is going to be a long job working out where it came from.
If you have an already well researched DNA and can place segments of DNA with known relatives then it is possible to go smaller (DNA painting comes in here)
One technique I use is to multiply the cM value by the number of SNPs, the bigger the score the more likely your going to find someone that touches your tree because both of you are likely to have known relatives to compare.
The thing for me is that the matches I want are the really small matches because I am back in the early 1800s and I have to find big matches to work their tree back to mine in some way.
Something I would recommend, as someone did already, is try to look for any match with given surnames off your tree. Obviously Smith might be a long shot.... However, also notice locations?
I have 12% Scandinavian genes as a surprise and one of the things I have been trying is collecting all the Scandinavian names, Johnson is the obvious name that is not in my tree. One thing that comes out of this is you will get many people from your 8000+ matches with the same surname - might be worth asking if a name recurs again and again. I have some notable links to people called Tucker with no idea where it comes from but it appears they have Scandinavian genes too!
When looking at possible matches I often use any combinations of these to make decisions about a letter of enquiry.
As the man said, "..it depends.." Make some fuzzy logic judgements on the items above and use your intuition too. It is isn't about science as such at this point?
One of the things lacking in DNA research is a how to guide on what to do with your results. Bits and pieces are given, but there is not an overall strategy as yet.
Hope this helps and please ask if I confused anyone?
Crowe, Crough, Croagh some Croak and Crow
Ireland mainly Tipperary / Limerick (where the name variations occurred most) but I have many all Ireland collections of data etc.

Offline Shay123

  • RootsChat Extra
  • **
  • Posts: 12
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: DNA what is the bare minimum in % for matches cms or segments ?
« Reply #10 on: Sunday 05 July 20 16:40 BST (UK) »
Such a rush...
I forgot to mention segments, more segments means a closer match even if the segments added together are smaller than one, single segment (in my experience).
Lastly, if you have a small match, look for others that also have the same match - you need to download the data for all your matches from your results dupplier. Make them into a spreadsheet and sort by Chromosome, Start point and End point in that order. Scanning the list will give you possibles and the names of contacts?
Shay
Crowe, Crough, Croagh some Croak and Crow
Ireland mainly Tipperary / Limerick (where the name variations occurred most) but I have many all Ireland collections of data etc.

Offline clayton bradley

  • RootsChat Veteran
  • *****
  • Posts: 980
  • Census information Crown Copyright, from www.natio
    • View Profile
Re: DNA what is the bare minimum in % for matches cms or segments ?
« Reply #11 on: Sunday 05 July 20 18:59 BST (UK) »
I have a number of people identified as having a common ancestor with me on Ancestry who have 7cMs. My 5xggfather Thomas Broadley, born abt 1733, married twice. I'm descended from his first wife and Ancestry picked up 3 first cousins (to each other) in America from Thomas's second wife.
Broadley (Lancs all dates and Halifax bef 1654)


Offline mona lisa

  • RootsChat Veteran
  • *****
  • Posts: 508
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: DNA what is the bare minimum in % for matches cms or segments ?
« Reply #12 on: Sunday 05 July 20 22:43 BST (UK) »
Also   with   the   7  cms    and  1 segment  thought maybe the 1 segment might be way too  insignificant ?   curious if the DNA Segments  lower  than say 5  segments are (hopeless)
Thank you Gadget   Hurworth    brigidmac     meelystar  melbaschmelba  Avm28   shay123    Clayton bradley
Ire Co Arm Seagoe Drumon JOYCE/HOUSTON John/Isabella ELIZ/ARMSTRONG /Wm Valentine + Hannah Johnstone MILLER aft1840 in Glasgow Scot dau Isabella 1847 + John M'D s/o Co DERRY Edw McD*V*TT//DAID + Eliz M'CL*SKY/M'SHANE/TEMPLETON/DUFF Ire = DOYLE,Jos + Mary M'GINITY/ m ca1840 = Can PEI Hope Riv/ DOWDS,Bridget b ca1819 d/o John DOWD + Mary WOODS ca1840 PEI Can UK IOM Wales, Barry ISL= REINE/RIIM = Adrianoff/ Michal =Arthur/Mary/Mildred / /MEREDITH/ TRAV*S/GARNETT/VON Stiegler/VON STIEGLITZ

Offline brigidmac

  • RootsChat Aristocrat
  • ******
  • Posts: 2,675
  • Computer incompetent but stiil trying
    • View Profile
Re: DNA what is the bare minimum in % for matches cms or segments ?
« Reply #13 on: Monday 06 July 20 13:59 BST (UK) »
Thanks for your explanations Shay I'm going to try out looking at some small matches linking to a larger match and look at names and locations. It will be a bit slap dash because I don't do spread sheets .so hope pseudo science works
Roberts,Fellman.Macdermid MCDERMID McDiarmid Gardner Jones ,Bloch,Irvine,Hallis Stevenson ,McKay

Offline brigidmac

  • RootsChat Aristocrat
  • ******
  • Posts: 2,675
  • Computer incompetent but stiil trying
    • View Profile
Re: DNA what is the bare minimum in % for matches cms or segments ?
« Reply #14 on: Monday 06 July 20 14:04 BST (UK) »
Here's my pseudo science
I have a match of 38cm across 3 segments
A couple on her tree b 1875 Russia
Have 5 daughters
Several links in name matches to  4 of their married names
Matches are between 6-15cm but it looks like some are several generations younger .
So I'm building up a provisional tree to see if I can find a connection to this couples parents ...maybe find an unknown great grandmother
Roberts,Fellman.Macdermid MCDERMID McDiarmid Gardner Jones ,Bloch,Irvine,Hallis Stevenson ,McKay

Offline Shay123

  • RootsChat Extra
  • **
  • Posts: 12
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: DNA what is the bare minimum in % for matches cms or segments ?
« Reply #15 on: Tuesday 07 July 20 06:22 BST (UK) »
Hello,
The important thing to remember about "does size matter" is that below 7cm od course you can get a match, however, for most people researching the distance in generations is likely to be large and after 6 or seven very difficult to find the connection in terms of time and effort unless two people have very extensive trees?
Many people can not take their tree back very far eg Irish or American / Australian researchers the numbers of generations in a tree may be limited. 5 generations in my Irish tree takes me back to the beginning of national records.
Of course too, it is important to remember, that by chance, a segment may be passed along for longer than 6 or 7 generations.
So the gist of it is to say that for most people when they start out with DNA is not to chase the smaller stuff unless the mix of evidence suggest the time spent researching it.
I am fortunate that I can mostly  spend as long as I want but even then, when people have only just started their tree, people do not reply, people have only one particular interest, people have a massive tree but do not want to divulge anything of it, it can still be difficult to find the time.
One last example and sorry for the personal example. I have two strong Scandanavian DNA links. My sister has our English line back to 1600s and we have connected with a know relative who has a 36% Scandi DNA connection and it is taking the three of us hours and hours of research over several weeks now and we still haven't unlocked it.
I know again that to find the clues to my own tree in Ireland I am working with small DNA connections but I have had to massively increase my tree and make new trees to track the connections between me and others. In fact it is becoming a one name study.
All DNA matches are potentially important but the question of size really related tot the amount of effort needed with small matches. "Triangulation" can help as brigidmac has explained.
Good luck!
Crowe, Crough, Croagh some Croak and Crow
Ireland mainly Tipperary / Limerick (where the name variations occurred most) but I have many all Ireland collections of data etc.