Why do the same DNA matches have differing centimorgans on GEDMatch vs 23andMe?

[Varg Vikernes 2]

Several of my 23andMe matches have also uploaded their results to GEDMatch, and consistently it seems that the 23andMe centimorgan count is lower (usually by around 25%) than the one I receive for the same person on GEDMatch. Why is this?

Here are some examples:

23andMe: 85CM
GEDMatch: 99.8

23andMe: 33CM
GEDMatch: 58.3

23andMe: 33CM
GEDMatch: 44CM

23andMe: 16CM
GEDMatch: 23.4CM


These are the same person on both 23andMe and GEDMatch, and the kits they've uploaded to GEDMatch come straight from 23andMe.

[Varg Vikernes 2]

And, more importantly, which one is more accurate?

[Mortimer]

Are you really varg vikernes?

[Abriekman]

Several of my 23andMe matches have also uploaded their results to GEDMatch, and consistently it seems that the 23andMe centimorgan count is lower (usually by around 25%) than the one I receive for the same person on GEDMatch. Why is this?

Here are some examples:

23andMe: 85CM
GEDMatch: 99.8

23andMe: 33CM
GEDMatch: 58.3

23andMe: 33CM
GEDMatch: 44CM

23andMe: 16CM
GEDMatch: 23.4CM


These are the same person on both 23andMe and GEDMatch, and the kits they've uploaded to GEDMatch come straight from 23andMe.

23andme does not count smaller segments. Gedmatch is more important for matches. Also can be that Gedmatch counts pileup ancient identical by descent segments. http://www.khazaria.com/genetics/excess-ibd-regions.html

[Varg Vikernes 2]

23andme does not count smaller segments. Gedmatch is more important for matches. Also can be that Gedmatch counts pileup ancient identical by descent segments. http://www.khazaria.com/genetics/excess-ibd-regions.html

If I do a 1 to 1 comparison between me and the match on both sites, it seems that 23andMe is missing some fairly large chunks (5CM to 10CM) so it doesn't seem to be counting up ancient DNA here.

[Leto]

Have you shared your Gedmatch results on the forum? I wonder what a half Maltese half Briton would score on calculators.

[Lemminkäinen]

Several of my 23andMe matches have also uploaded their results to GEDMatch, and consistently it seems that the 23andMe centimorgan count is lower (usually by around 25%) than the one I receive for the same person on GEDMatch. Why is this?

Here are some examples:

23andMe: 85CM
GEDMatch: 99.8

23andMe: 33CM
GEDMatch: 58.3

23andMe: 33CM
GEDMatch: 44CM

23andMe: 16CM
GEDMatch: 23.4CM


These are the same person on both 23andMe and GEDMatch, and the kits they've uploaded to GEDMatch come straight from 23andMe.

It depends on how long mismatching gaps are accepted.

[Varg Vikernes 2]

It depends on how long mismatching gaps are accepted.

Some of additional matching chunks GEDMatch identifies are on different chromosomes though, so its not just a result of mismatching gaps being accepted.

[Lemminkäinen]

Some of additional matching chunks GEDMatch identifies are on different chromosomes though, so its not just a result of mismatching gaps being accepted.

Can be. Think twice. Key words are max mismatching gap and min chunk length.

[Lemminkäinen]

Is the mismatch counted in basepairs, cMs or SNPs. Is the mismatch length the total length inside each chunk or only inside each mismatch. How many mismatches is allowed per chunk.

[Abriekman]

If I do a 1 to 1 comparison between me and the match on both sites, it seems that 23andMe is missing some fairly large chunks (5CM to 10CM) so it doesn't seem to be counting up ancient DNA here.

You can even share 20 cM identical by descent segment with somebody, so it can be easily false match. Check if segments you match are here http://www.khazaria.com/genetics/exc...d-regions.html