Announcement

Collapse
No announcement yet.

Autosomal DNA Skipping Generations

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Autosomal DNA Skipping Generations

    Why do I have several autosomal matches with people whom, I can identify where in the trees our ancestors connect (on my father's side) and yet, these people doesn't share a match with my father, his sisters, my cousins on that side. How does autosomal information disappear on a generation but pop up in the next?

  • #2
    Originally posted by cjm View Post
    Why do I have several autosomal matches with people whom, I can identify where in the trees our ancestors connect (on my father's side) and yet, these people doesn't share a match with my father, his sisters, my cousins on that side. How does autosomal information disappear on a generation but pop up in the next?
    Matching is based on set criteria.
    Here at FTDNA to be declared a match one must have a longest segment of at least 7cM/500SNPs and share at least 20cM in total DNA.
    Failing to meet that criteria, you will also be shown any match that shares under 20cM Total DNA if that match has a longest segment of 9cM or greater.

    Many times due to this, match may not show up in parent but does in child (matches where longest segment is under 9cM).
    ie)Behind the scene, Matching Algorithm shows parent and match with a longest segment of 8cM and total shared DNA as 18cM, they will be rejected as showing up on each others match lists since longest segment is under 9cM and total DNA Shared is under 20cM

    It shows in child due to the total of all segments 1cM/500SNPs or greater just happen to push total DNA shared over the 20cM mark. Note most segments under 7cM are false positive segments.

    If you are able to check match with parent at Gedmatch, you will most likely see the same longest segment in parent in question.

    Most testing companies have a set criteria for determining a match.
    Gedmatch only uses the 7cM/500 SNP criteria.
    It is left up to user to decide on validity as well they allow you to lower thresholds.
    Last edited by prairielad; 10th June 2018, 11:29 PM.

    Comment


    • #3
      While some physical characteristics (as a result of gene expression, such as "recessive" traits) might be observed to skip a generation, DNA itself does NOT skip generations: You have to get your DNA from your parents, not directly from your grandparents.

      The algorithms used to detect "matching segments" are not perfect and can be fooled, at least occasionally, resulting in apparently inconsistent results such as those reported in this thread. Sometimes the misbehaving segments are fairly long, on the order of 10 cM (I have a group of "matches" on GEDmatch that don't triangulate consistently with each other on chromosome 22). Exactly why this happens with any particular algorithm is hard to discover, because I don't have access to the raw data for the matches, nor do I have access to the complete matching algorithm. Nevertheless, MOST segments are well behaved, clearly falling into paternal or maternal triangulation groups, for segments about 10 cM or longer. When I encounter an isolated segment that doesn't behave logically, in a situation where the family tree is not at issue, I just put it aside, in hopes that additional matches will eventually turn up that will clarify the situation.

      Comment

      Working...
      X