Archiver > GENEALOGY-DNA > 2010-02 > 1266517185

From: Michael Maddi <>
Subject: [DNA] Relative Finder predicted 3rd cousin is actually 2nd cousin
Date: Thu, 18 Feb 2010 10:19:45 -0800 (PST)

I have now been in contact with my closest predicted Relative Finder cousin at 23andMe. She is predicted to be a 3rd cousin (range - 2nd-4th), with whom I share 1.47% of DNA in 5 segments.

She turns out to be a second cousin, with whom I haven't been in contact before, although I have been in contact with other 2nd cousins, who are first cousins to her. We all share common ancestors in our great-grandparents, Nunzio Maddi and Rosalia Cangialosi.

This is really amazing! Who would think that two great-grandchildren of Nunzio Maddi and Rosalia Cangialosi would both test at 23andMe, which only has about 30,000 customers?

It's interesting that 23andMe have predicted us as 3rd cousins, with a possible range of 2nd-4th cousins. Most people are finding that their Relative Finder cousins are more distant than predicted, yet we're closer than predicted.

Here's a chart showing the common DNA segments we have:

Chromosome Start point End point Genetic distance # SNPs
4 145000000 182000000 42 cM 6632
7 75000000 105000000 31 cM 5967
10 16000000 24000000 8.0 cM 1696
21 36000000 41000000 10.0 cM 1347
22 43000000 49000000 18.5 cM 1981

The shared segments on chromosomes 4 and 7 are huge. I'm wondering if we have more common ancestors than our great-grandparents to share so many segments. Most Relative Finder cousins share 1 or 2 segments and we share 5.

I notice that her father has an Italian surname, Ferrazano, and wonder if we might also have common ancestors on that side as well. Using the World Names Profiler search engine at, there are no Ferrazanos found in Italy. However, there are Ferrazzanos found in a couple of Italian regions where my mother had ancestors. These are Puglia and Basilicata, although the towns where the Ferrazzano surname occurs are not that close to the towns of my mother's ancestors. So I think that our shared segments are probably solely from either Nunzio or Rosalia or both.

Connecting up with her has been a real find. Since she's in the strict maternal line of my great-grandmother (her daughter was my second cousin's maternal grandmother), I now know that Rosalia's mitochondrial haplogroup was U1a. Also, since my original reason for getting involved in DNA testing was to see if I could find out the name of Nunzio Maddi's real father, it's possible that triangulating our Relative Finder lists, we may be able to determine which of our shared segments came from Nunzio and use that to see who also shares that segment(s) with us.

That could offer a clue to the surname of my great-grandfather's birth father. (He was abandoned at birth, so Maddi is not the true surname of my paternal line.) I already have a clue from yDNA testing that the true surname may be Salerno, since I match 43/43 in the database with someone named Salerno with ancestors from Palermo or the area, about 25 miles from where my great-grandfather lived. Now if I find a cousin with Salerno in his family tree in either 23andMe's database or in FTDNA's Family Finder database, once it gets off the ground, that will help me in my goal of knocking down this brick wall.

I don't know if my good luck will continue, but one lesson here is to put your DNA results in as many databases as possible and, if you can afford it, test with multiple companies.

Mike Maddi

This thread: