Half-1st cousin?

[prairielad]

My Uncle and his 1st cousins longest block is 54.22 cM, but shares a total of 792.82.

David Pikes utility lists segment as 42.50 Mb

[hansonrf]

I could test the sisters but even if they show in the half sister range will that prove any more than this result?

The sisters will be very clear; around 1900 total match all 1/2 IBD for half-sibs and 2700 total, 1900 half IBD and 800 FULL IBD, for full sibs.

[MisterPC]

For 1st cousins, the "shortest longest" segment I've seen referenced is 50 cM.

Was the total 530cM and longest 50cM for the same relation? How did you manage to confirm this was a full 1st cousin and not a half cousin or some other similar relation?

[hfp43]

Half 1st-Cousin

The largest segment (50 cM) and the "most total" (530 cM) were both taken from an analysis done by an autosomal expert (which I am not) and are probably not from the same match. Both numbers are extreme and each is a low probability. You haven't asked us directly, but I think in this (your) case an NPE is a reasonable hypothesis. I would not jump to that conclusion, but I would certainly want to test it. A key question is whether you can do that without freaking somebody out.

[Serket]

My first cousin tested on Ancestry and so we used Gedmatch to compare our results and we got Shared CM of 596.4 and the Longest Block is 48.6.

According to the Autosomal DNA statistics by ISOGG, 23andMe's Relative Finder claims that 1st cousins can be in the range of 7.31% to 13.8% and their page says to divide your Shared CM results by 71.6 to get a percentage. For my cousin's Gedmatch results that would work out to 8.33% which is on the low end, but still within range.

When I run your number of 503.33 it comes out to 7.03% which is too low for that range. However you said something about the sisters being half-siblings. Do you already know that your mother and your cousin's mother are half-sisters rather than full sisters?